Last Year

Article
Beigpour, S., Shekhar, S., Mansouryar, M., Myszkowski, K., and Seidel, H.-P. 2018. Light-Field Appearance Editing Based on Intrinsic Decomposition. Journal of Perceptual Imaging1, 1.
Export
BibTeX
@article{Beigpour2018, TITLE = {Light-Field Appearance Editing Based on Intrinsic Decomposition}, AUTHOR = {Beigpour, Shida and Shekhar, Sumit and Mansouryar, Mohsen and Myszkowski, Karol and Seidel, Hans-Peter}, LANGUAGE = {eng}, DOI = {10.2352/J.Percept.Imaging.2018.1.1.010502}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, JOURNAL = {Journal of Perceptual Imaging}, VOLUME = {1}, NUMBER = {1}, PAGES = {1--15}, EID = {10502}, }
Endnote
%0 Journal Article %A Beigpour, Shida %A Shekhar, Sumit %A Mansouryar, Mohsen %A Myszkowski, Karol %A Seidel, Hans-Peter %+ External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T Light-Field Appearance Editing Based on Intrinsic Decomposition : %G eng %U http://hdl.handle.net/21.11116/0000-0001-5F88-C %R 10.2352/J.Percept.Imaging.2018.1.1.010502 %7 2018 %D 2018 %J Journal of Perceptual Imaging %O JPI %V 1 %N 1 %& 1 %P 1 - 15 %Z sequence number: 10502
Chen, R., Gotsman, C., and Hormann, K. 2018a. Path Planning with Divergence-Based Distance Functions. Computer Aided Geometric Design66.
Export
BibTeX
@article{Chen_CAGD2018, TITLE = {Path Planning with Divergence-Based Distance Functions}, AUTHOR = {Chen, Renjie and Gotsman, Craig and Hormann, Kai}, LANGUAGE = {eng}, ISSN = {0167-8396}, DOI = {10.1016/j.cagd.2018.09.002}, PUBLISHER = {North-Holland}, ADDRESS = {Amsterdam}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {Computer Aided Geometric Design}, VOLUME = {66}, PAGES = {52--74}, }
Endnote
%0 Journal Article %A Chen, Renjie %A Gotsman, Craig %A Hormann, Kai %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations %T Path Planning with Divergence-Based Distance Functions : %G eng %U http://hdl.handle.net/21.11116/0000-0002-72C1-3 %R 10.1016/j.cagd.2018.09.002 %7 2018 %D 2018 %J Computer Aided Geometric Design %V 66 %& 52 %P 52 - 74 %I North-Holland %C Amsterdam %@ false
Chen, R., Gotsman, C., and Hormann, K. 2018b. Efficient Path Generation with Reduced Coordinates. Computer Graphics Forum (Proc. Eurographics Symposium on Geometric Processing 2018)37, 5.
Export
BibTeX
@article{ChenSGP2018, TITLE = {Efficient Path Generation with Reduced Coordinates}, AUTHOR = {Chen, Renjie and Gotsman, Craig and Hormann, Kai}, LANGUAGE = {eng}, ISSN = {0167-7055}, DOI = {10.1111/cgf.13489}, PUBLISHER = {Wiley-Blackwell}, ADDRESS = {Chichester}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {Computer Graphics Forum (Proc. Eurographics Symposium on Geometric Processing)}, VOLUME = {37}, NUMBER = {5}, PAGES = {37--48}, BOOKTITLE = {Symposium on Geometry Processing 2018 (Eurographics Symposium on Geometric Processing 2018)}, EDITOR = {Ju, Tao and Vaxman, Amir}, }
Endnote
%0 Journal Article %A Chen, Renjie %A Gotsman, Craig %A Hormann, Kai %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations %T Efficient Path Generation with Reduced Coordinates : %G eng %U http://hdl.handle.net/21.11116/0000-0001-E6D6-A %R 10.1111/cgf.13489 %7 2018 %D 2018 %J Computer Graphics Forum %V 37 %N 5 %& 37 %P 37 - 48 %I Wiley-Blackwell %C Chichester %@ false %B Symposium on Geometry Processing 2018 %O Paris, France, July 7 – 11, 2018 SGP 2018 Eurographics Symposium on Geometric Processing 2018
Du, X., Liu, X., Yan, D.-M., Jiang, C., Ye, J., and Zhang, H. 2018. Field-Aligned Isotropic Surface Remeshing. Computer Graphics Forum37, 6.
Export
BibTeX
@article{DuCGF2018, TITLE = {Field-Aligned Isotropic Surface Remeshing}, AUTHOR = {Du, Xingyi and Liu, Xiaohan and Yan, Dong-Ming and Jiang, Caigui and Ye, Juntao and Zhang, Hui}, LANGUAGE = {eng}, ISSN = {0167-7055}, DOI = {10.1111/cgf.13329}, PUBLISHER = {Blackwell-Wiley}, ADDRESS = {Oxford}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {Computer Graphics Forum}, VOLUME = {37}, NUMBER = {6}, PAGES = {343--357}, }
Endnote
%0 Journal Article %A Du, Xingyi %A Liu, Xiaohan %A Yan, Dong-Ming %A Jiang, Caigui %A Ye, Juntao %A Zhang, Hui %+ External Organizations External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations %T Field-Aligned Isotropic Surface Remeshing : %G eng %U http://hdl.handle.net/21.11116/0000-0001-E209-6 %R 10.1111/cgf.13329 %7 2018 %D 2018 %J Computer Graphics Forum %O Computer Graphics Forum : journal of the European Association for Computer Graphics Comput. Graph. Forum %V 37 %N 6 %& 343 %P 343 - 357 %I Blackwell-Wiley %C Oxford %@ false
Golla, B., Seidel, H.-P., and Chen, R. 2018. Piecewise Linear Mapping Optimization Based on the Complex View. Computer Graphics Forum (Proc. Pacific Graphics 2018)37, 7.
Export
BibTeX
@article{Golla_PG2018, TITLE = {Piecewise Linear Mapping Optimization Based on the Complex View}, AUTHOR = {Golla, Bj{\"o}rn and Seidel, Hans-Peter and Chen, Renjie}, LANGUAGE = {eng}, ISSN = {1467-8659}, DOI = {10.1111/cgf.13563}, PUBLISHER = {Wiley-Blackwell}, ADDRESS = {Oxford, UK}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {Computer Graphics Forum (Proc. Pacific Graphics)}, VOLUME = {37}, NUMBER = {7}, PAGES = {233--243}, BOOKTITLE = {The 26th Pacific Conference on Computer Graphics and Applications (Pacific Graphics 2018)}, }
Endnote
%0 Journal Article %A Golla, Björn %A Seidel, Hans-Peter %A Chen, Renjie %+ External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T Piecewise Linear Mapping Optimization Based on the Complex View : %G eng %U http://hdl.handle.net/21.11116/0000-0002-72CD-7 %R 10.1111/cgf.13563 %7 2018 %D 2018 %J Computer Graphics Forum %V 37 %N 7 %& 233 %P 233 - 243 %I Wiley-Blackwell %C Oxford, UK %@ false %B The 26th Pacific Conference on Computer Graphics and Applications %O Pacific Graphics 2018 PG 2018 Hong Kong, 8-11 October 2018
Hladký, J. and Ďurikovič, R. 2018. Fire Simulation in 3D Computer Animation with Turbulence Dynamics including Fire Separation and Profile Modeling. International Journal of Networking and Computing8, 2.
Export
BibTeX
@article{DBLP:journals/ijnc/HladkyD18, TITLE = {Fire Simulation in {3D} Computer Animation with Turbulence Dynamics including Fire Separation and Profile Modeling}, AUTHOR = {Hladk{\'y}, Jozef and {\v D}urikovi{\v c}, Roman}, LANGUAGE = {eng}, ISSN = {2185-2847}, URL = {http://www.ijnc.org/index.php/ijnc/article/view/180}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, JOURNAL = {International Journal of Networking and Computing}, VOLUME = {8}, NUMBER = {2}, PAGES = {186--204}, }
Endnote
%0 Journal Article %A Hladký, Jozef %A Ďurikovič, Roman %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations %T Fire Simulation in 3D Computer Animation with Turbulence Dynamics including Fire Separation and Profile Modeling : %G eng %U http://hdl.handle.net/21.11116/0000-0002-F652-C %U http://www.ijnc.org/index.php/ijnc/article/view/180 %7 2018 %D 2018 %J International Journal of Networking and Computing %V 8 %N 2 %& 186 %P 186 - 204 %@ false
Kenzel, M., Kerbl, B., Schmalstieg, D., and Steinberger, M. 2018. A High-Performance Software Graphics Pipeline Architecture for the GPU. ACM Transactions on Graphics (Proc. ACM SIGGRAPH 2018)37, 4.
Export
BibTeX
@article{Kenzel_SIGGRAPH2018, TITLE = {A High-Performance Software Graphics Pipeline Architecture for the {GPU}}, AUTHOR = {Kenzel, Michael and Kerbl, Bernhard and Schmalstieg, Dieter and Steinberger, Markus}, LANGUAGE = {eng}, ISSN = {0730-0301}, DOI = {10.1145/3197517.3201374}, PUBLISHER = {ACM}, ADDRESS = {New York, NY}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {ACM Transactions on Graphics (Proc. ACM SIGGRAPH)}, VOLUME = {37}, NUMBER = {4}, PAGES = {1--15}, EID = {140}, BOOKTITLE = {Proceedings of ACM SIGGRAPH 2018}, }
Endnote
%0 Journal Article %A Kenzel, Michael %A Kerbl, Bernhard %A Schmalstieg, Dieter %A Steinberger, Markus %+ External Organizations External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T A High-Performance Software Graphics Pipeline Architecture for the GPU : %G eng %U http://hdl.handle.net/21.11116/0000-0002-72E1-F %R 10.1145/3197517.3201374 %7 2018 %D 2018 %J ACM Transactions on Graphics %V 37 %N 4 %& 1 %P 1 - 15 %Z sequence number: 140 %I ACM %C New York, NY %@ false %B Proceedings of ACM SIGGRAPH 2018 %O ACM SIGGRAPH 2018 Vancouver, Canada , 12 - 16 August
Kim, H., Garrido, P., Tewari, A., et al. 2018a. Deep Video Portraits. ACM Transactions on Graphics (Proc. ACM SIGGRAPH 2018)37, 4.
Export
BibTeX
@article{Kim_SIGGRAPH2018, TITLE = {Deep Video Portraits}, AUTHOR = {Kim, Hyeongwoo and Garrido, Pablo and Tewari, Ayush and Xu, Weipeng and Thies, Justus and Nie{\ss}ner, Matthias and P{\'e}rez, Patrick and Richardt, Christian and Zollh{\"o}fer, Michael and Theobalt, Christian}, LANGUAGE = {eng}, ISSN = {0730-0301}, DOI = {10.1145/3197517.3201283}, PUBLISHER = {ACM}, ADDRESS = {New York, NY}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {ACM Transactions on Graphics (Proc. ACM SIGGRAPH)}, VOLUME = {37}, NUMBER = {4}, PAGES = {1--14}, EID = {163}, BOOKTITLE = {Proceedings of ACM SIGGRAPH 2018}, }
Endnote
%0 Journal Article %A Kim, Hyeongwoo %A Garrido, Pablo %A Tewari, Ayush %A Xu, Weipeng %A Thies, Justus %A Nießner, Matthias %A Pérez, Patrick %A Richardt, Christian %A Zollhöfer, Michael %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations External Organizations External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T Deep Video Portraits : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5D8D-8 %R 10.1145/3197517.3201283 %7 2018 %D 2018 %J ACM Transactions on Graphics %V 37 %N 4 %& 1 %P 1 - 14 %Z sequence number: 163 %I ACM %C New York, NY %@ false %B Proceedings of ACM SIGGRAPH 2018 %O ACM SIGGRAPH 2018 Vancouver, Canada , 12 - 16 August
Lee, H. and Didyk, P. 2018. Real-time Apparent Resolution Enhancement for Head-mounted Displays. Proceedings of the ACM on Computer Graphics and Interactive Techniques1, 1.
Export
BibTeX
@article{Lee:2018:RAR:3242771.3203202, TITLE = {Real-time Apparent Resolution Enhancement for Head-mounted Displays}, AUTHOR = {Lee, Haebom and Didyk, Piotr}, LANGUAGE = {eng}, ISSN = {2577-6193}, DOI = {10.1145/3203202}, PUBLISHER = {ACM}, ADDRESS = {New York, NY}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, JOURNAL = {Proceedings of the ACM on Computer Graphics and Interactive Techniques}, VOLUME = {1}, NUMBER = {1}, PAGES = {1--15}, EID = {19}, }
Endnote
%0 Journal Article %A Lee, Haebom %A Didyk, Piotr %+ External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T Real-time Apparent Resolution Enhancement for Head-mounted Displays : %G eng %U http://hdl.handle.net/21.11116/0000-0002-DB97-D %R 10.1145/3203202 %7 2018 %D 2018 %J Proceedings of the ACM on Computer Graphics and Interactive Techniques %V 1 %N 1 %& 1 %P 1 - 15 %Z sequence number: 19 %I ACM %C New York, NY %@ false
Leimkühler, T., Seidel, H.-P., and Ritschel, T. 2018a. Laplacian Kernel Splatting for Efficient Depth-of-field and Motion Blur Synthesis or Reconstruction. ACM Transactions on Graphics (Proc. ACM SIGGRAPH 2018)37, 4.
Export
BibTeX
@article{LeimkuehlerSIGGRAPH2018, TITLE = {Laplacian Kernel Splatting for Efficient Depth-of-field and Motion Blur Synthesis or Reconstruction}, AUTHOR = {Leimk{\"u}hler, Thomas and Seidel, Hans-Peter and Ritschel, Tobias}, LANGUAGE = {eng}, ISSN = {0730-0301}, DOI = {10.1145/3197517.3201379}, PUBLISHER = {ACM}, ADDRESS = {New York, NY}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {ACM Transactions on Graphics (Proc. ACM SIGGRAPH)}, VOLUME = {37}, NUMBER = {4}, PAGES = {1--11}, EID = {55}, BOOKTITLE = {Proceedings of ACM SIGGRAPH 2018}, }
Endnote
%0 Journal Article %A Leimkühler, Thomas %A Seidel, Hans-Peter %A Ritschel, Tobias %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations %T Laplacian Kernel Splatting for Efficient Depth-of-field and Motion Blur Synthesis or Reconstruction : %G eng %U http://hdl.handle.net/21.11116/0000-0002-0630-1 %R 10.1145/3197517.3201379 %7 2018 %D 2018 %J ACM Transactions on Graphics %V 37 %N 4 %& 1 %P 1 - 11 %Z sequence number: 55 %I ACM %C New York, NY %@ false %B Proceedings of ACM SIGGRAPH 2018 %O ACM SIGGRAPH 2018 Vancouver, Canada , 12 - 16 August
Leimkühler, T., Kellnhofer, P., Ritschel, T., Myszkowski, K., and Seidel, H.-P. 2018b. Perceptual Real-Time 2D-to-3D Conversion Using Cue Fusion. IEEE Transactions on Visualization and Computer Graphics24, 6.
Export
BibTeX
@article{Leimkuehler2018, TITLE = {Perceptual real-time {2D}-to-{3D} conversion using cue fusion}, AUTHOR = {Leimk{\"u}hler, Thomas and Kellnhofer, Petr and Ritschel, Tobias and Myszkowski, Karol and Seidel, Hans-Peter}, LANGUAGE = {eng}, ISSN = {1077-2626}, DOI = {10.1109/TVCG.2017.2703612}, PUBLISHER = {IEEE Computer Society}, ADDRESS = {New York, NY}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {IEEE Transactions on Visualization and Computer Graphics}, VOLUME = {24}, NUMBER = {6}, PAGES = {2037--2050}, }
Endnote
%0 Journal Article %A Leimkühler, Thomas %A Kellnhofer, Petr %A Ritschel, Tobias %A Myszkowski, Karol %A Seidel, Hans-Peter %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T Perceptual Real-Time 2D-to-3D Conversion Using Cue Fusion : %G eng %U http://hdl.handle.net/21.11116/0000-0001-409A-9 %R 10.1109/TVCG.2017.2703612 %7 2018 %D 2018 %J IEEE Transactions on Visualization and Computer Graphics %V 24 %N 6 %& 2037 %P 2037 - 2050 %I IEEE Computer Society %C New York, NY %@ false
Liu, L., Chen, N., Ceylan, D., Theobalt, C., Wang, W., and Mitra, N.J. 2018a. CurveFusion: Reconstructing thin Structures from RGBD Sequences. ACM Transactions on Graphics37, 6.
Export
BibTeX
@article{Liu:2018:CvF, TITLE = {{CurveFusion}: reconstructing thin structures from {RGBD} sequences}, AUTHOR = {Liu, Lingjie and Chen, Nenglun and Ceylan, Duygu and Theobalt, Christian and Wang, Wenping and Mitra, Niloy J.}, LANGUAGE = {eng}, ISSN = {0730-0301}, DOI = {10.1145/3272127.3275097}, PUBLISHER = {ACM}, ADDRESS = {New York, NY}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {ACM Transactions on Graphics}, VOLUME = {37}, NUMBER = {6}, PAGES = {1--12}, EID = {2018}, }
Endnote
%0 Journal Article %A Liu, Lingjie %A Chen, Nenglun %A Ceylan, Duygu %A Theobalt, Christian %A Wang, Wenping %A Mitra, Niloy J. %+ External Organizations External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations %T CurveFusion: Reconstructing thin Structures from RGBD Sequences : %G eng %U http://hdl.handle.net/21.11116/0000-0002-E6F2-9 %R 10.1145/3272127.3275097 %7 2018 %D 2018 %J ACM Transactions on Graphics %V 37 %N 6 %& 1 %P 1 - 12 %Z sequence number: 2018 %I ACM %C New York, NY %@ false
Piovarči, M., Levin, D.I.W., Kaufman, D.M., and Didyk, P. 2018a. Perception-Aware Modeling and Fabrication of Digital Drawing Tools. ACM Transactions on Graphics (Proc. ACM SIGGRAPH 2018)37, 4.
Export
BibTeX
@article{Piovarci_SIGGRAPH2018, TITLE = {Perception-Aware Modeling and Fabrication of Digital Drawing Tools}, AUTHOR = {Piovar{\v c}i, Michal and Levin, David I. W. and Kaufman, Danny M. and Didyk, Piotr}, LANGUAGE = {eng}, ISSN = {0730-0301}, DOI = {10.1145/3197517.3201322}, PUBLISHER = {ACM}, ADDRESS = {New York, NY}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {ACM Transactions on Graphics (Proc. ACM SIGGRAPH)}, VOLUME = {37}, NUMBER = {4}, PAGES = {1--15}, EID = {123}, BOOKTITLE = {Proceedings of ACM SIGGRAPH 2018}, }
Endnote
%0 Journal Article %A Piovarči, Michal %A Levin, David I. W. %A Kaufman, Danny M. %A Didyk, Piotr %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T Perception-Aware Modeling and Fabrication of Digital Drawing Tools : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5D7D-B %R 10.1145/3197517.3201322 %7 2018 %D 2018 %J ACM Transactions on Graphics %V 37 %N 4 %& 1 %P 1 - 15 %Z sequence number: 123 %I ACM %C New York, NY %@ false %B Proceedings of ACM SIGGRAPH 2018 %O ACM SIGGRAPH 2018 Vancouver, Canada , 12 - 16 August
Piovarči, M., Wessely, M., Jagielski, M., Alexa, M., Matusik, W., and Didyk, P. 2018b. Design and Analysis of Directional Front Projection Screens. Computers and Graphics74.
Export
BibTeX
@article{Piovarci_2018, TITLE = {Design and Analysis of Directional Front Projection Screens}, AUTHOR = {Piovar{\v c}i, Michal and Wessely, Michael and Jagielski, Michal and Alexa, Marc and Matusik, Wojciech and Didyk, Piotr}, LANGUAGE = {eng}, ISSN = {0097-8493}, DOI = {10.1016/j.cag.2018.05.010}, PUBLISHER = {Pergamon}, ADDRESS = {New York}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {Computers and Graphics}, VOLUME = {74}, PAGES = {213--224}, }
Endnote
%0 Journal Article %A Piovarči, Michal %A Wessely, Michael %A Jagielski, Michal %A Alexa, Marc %A Matusik, Wojciech %A Didyk, Piotr %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T Design and Analysis of Directional Front Projection Screens : %G eng %U http://hdl.handle.net/21.11116/0000-0002-16D9-1 %R 10.1016/j.cag.2018.05.010 %7 2018 %D 2018 %J Computers and Graphics %V 74 %& 213 %P 213 - 224 %I Pergamon %C New York %@ false
Shi, L., Babaei, V., Kim, C., et al. 2018. Deep Multispectral Painting Reproduction via Multi-Layer, Custom-Ink Printing. ACM Transactions on Graphics (Proc. ACM SIGGRAPH Asia 2018)37, 6.
Export
BibTeX
@article{Shi2018, TITLE = {Deep Multispectral Painting Reproduction via Multi-Layer, Custom-Ink Printing}, AUTHOR = {Shi, Liang and Babaei, Vahid and Kim, Chagil and Foshey, Michael and Hu, Yuanming and Sitthi-Amorn, Pitchaya and Rusinkiewicz, Szymon and Matusik, Wojciech}, LANGUAGE = {eng}, ISSN = {0730-0301}, ISBN = {978-1-4503-6008-1}, DOI = {10.1145/3272127.3275057}, PUBLISHER = {ACM}, ADDRESS = {New York, NY}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {ACM Transactions on Graphics (Proc. ACM SIGGRAPH Asia)}, VOLUME = {37}, NUMBER = {6}, EID = {271}, BOOKTITLE = {Proceedings of ACM SIGGRAPH Asia 2018}, }
Endnote
%0 Journal Article %A Shi, Liang %A Babaei, Vahid %A Kim, Chagil %A Foshey, Michael %A Hu, Yuanming %A Sitthi-Amorn, Pitchaya %A Rusinkiewicz, Szymon %A Matusik, Wojciech %+ External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations External Organizations External Organizations External Organizations External Organizations %T Deep Multispectral Painting Reproduction via Multi-Layer, Custom-Ink Printing : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5C93-1 %R 10.1145/3272127.3275057 %7 2018 %D 2018 %J ACM Transactions on Graphics %V 37 %N 6 %Z sequence number: 271 %I ACM %C New York, NY %@ false %B Proceedings of ACM SIGGRAPH Asia 2018 %O ACM SIGGRAPH Asia 2018 Tokyo, Japan, December 04 - 07, 2018 SA'18 SA 2018 %@ 978-1-4503-6008-1
Singh, G., Subr, K., Coeurjolly, D., Ostromoukhov, V., and Jarosz, W. Analyzing and Improving Correlated Monte Carlo Importance Sampling of Discontinuous functions. Computer Graphics Forum.
(Accepted/in press)
Export
BibTeX
@article{SinghCGF2018, TITLE = {Analyzing and Improving Correlated {Monte Carlo} Importance Sampling of Discontinuous functions}, AUTHOR = {Singh, Gurprit and Subr, Kartic and Coeurjolly, David and Ostromoukhov, Victor and Jarosz, Wojciech}, LANGUAGE = {eng}, ISSN = {0167-7055}, PUBLISHER = {Blackwell-Wiley}, ADDRESS = {Oxford}, YEAR = {2018}, PUBLREMARK = {Accepted}, MARGINALMARK = {$\bullet$}, JOURNAL = {Computer Graphics Forum}, }
Endnote
%0 Journal Article %A Singh, Gurprit %A Subr, Kartic %A Coeurjolly, David %A Ostromoukhov, Victor %A Jarosz, Wojciech %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations External Organizations External Organizations %T Analyzing and Improving Correlated Monte Carlo Importance Sampling of Discontinuous functions : %G eng %U http://hdl.handle.net/21.11116/0000-0002-9F87-3 %D 2018 %J Computer Graphics Forum %O Computer Graphics Forum : journal of the European Association for Computer Graphics Comput. Graph. Forum %I Blackwell-Wiley %C Oxford %@ false
Tewari, A., Zollhöfer, M., Bernard, F., et al. 2018a. High-Fidelity Monocular Face Reconstruction based on an Unsupervised Model-based Face Autoencoder. IEEE Transactions on Pattern Analysis and Machine Intelligence.
Export
BibTeX
@article{8496850, TITLE = {High-Fidelity Monocular Face Reconstruction based on an Unsupervised Model-based Face Autoencoder}, AUTHOR = {Tewari, Ayush and Zollh{\"o}fer, Michael and Bernard, Florian and Garrido, Pablo and Kim, Hyeongwoo and P{\'e}rez, Patrick and Theobalt, Christian}, LANGUAGE = {eng}, ISSN = {0162-8828}, DOI = {10.1109/TPAMI.2018.2876842}, PUBLISHER = {IEEE}, ADDRESS = {Piscataway, NJ}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, JOURNAL = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, }
Endnote
%0 Journal Article %A Tewari, Ayush %A Zollhöfer, Michael %A Bernard, Florian %A Garrido, Pablo %A Kim, Hyeongwoo %A Pérez, Patrick %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T High-Fidelity Monocular Face Reconstruction based on an Unsupervised Model-based Face Autoencoder : %G eng %U http://hdl.handle.net/21.11116/0000-0002-EF5B-C %R 10.1109/TPAMI.2018.2876842 %7 2018 %D 2018 %J IEEE Transactions on Pattern Analysis and Machine Intelligence %O IEEE Trans. Pattern Anal. Mach. Intell. %I IEEE %C Piscataway, NJ %@ false
Thies, J., Zollhöfer, M., Theobalt, C., Stamminger, M., and Nießner, M. 2018a. HeadOn: Real-time Reenactment of Human Portrait Videos. ACM Transactions on Graphics (Proc. ACM SIGGRAPH 2018)37, 4.
Export
BibTeX
@article{Thies_SIGGRAPH2018, TITLE = {{HeadOn}: {R}eal-time Reenactment of Human Portrait Videos}, AUTHOR = {Thies, Justus and Zollh{\"o}fer, Michael and Theobalt, Christian and Stamminger, Marc and Nie{\ss}ner, Matthias}, LANGUAGE = {eng}, ISSN = {0730-0301}, DOI = {10.1145/3197517.3201350}, PUBLISHER = {ACM}, ADDRESS = {New York, NY}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {ACM Transactions on Graphics (Proc. ACM SIGGRAPH)}, VOLUME = {37}, NUMBER = {4}, PAGES = {1--13}, EID = {164}, BOOKTITLE = {Proceedings of ACM SIGGRAPH 2018}, }
Endnote
%0 Journal Article %A Thies, Justus %A Zollhöfer, Michael %A Theobalt, Christian %A Stamminger, Marc %A Nießner, Matthias %+ External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations %T HeadOn: Real-time Reenactment of Human Portrait Videos : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5D91-2 %R 10.1145/3197517.3201350 %7 2018 %D 2018 %J ACM Transactions on Graphics %V 37 %N 4 %& 1 %P 1 - 13 %Z sequence number: 164 %I ACM %C New York, NY %@ false %B Proceedings of ACM SIGGRAPH 2018 %O ACM SIGGRAPH 2018 Vancouver, Canada , 12 - 16 August
Thies, J., Zollhöfer, M., Stamminger, M., Theobalt, C., and Nießner, M. 2018b. FaceVR: Real-Time Gaze-Aware Facial Reenactment in Virtual Reality. ACM Transactions on Graphics37, 2.
Export
BibTeX
@article{thies18FaceVR, TITLE = {{FaceVR}: Real-Time Gaze-Aware Facial Reenactment in Virtual Reality}, AUTHOR = {Thies, Justus and Zollh{\"o}fer, Michael and Stamminger, Marc and Theobalt, Christian and Nie{\ss}ner, Matthias}, LANGUAGE = {eng}, ISSN = {0730-0301}, DOI = {10.1145/3182644}, PUBLISHER = {ACM}, ADDRESS = {New York, NY}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {ACM Transactions on Graphics}, VOLUME = {37}, NUMBER = {2}, PAGES = {1--15}, EID = {25}, }
Endnote
%0 Journal Article %A Thies, Justus %A Zollhöfer, Michael %A Stamminger, Marc %A Theobalt, Christian %A Nießner, Matthias %+ External Organizations External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations %T FaceVR: Real-Time Gaze-Aware Facial Reenactment in Virtual Reality : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5D96-D %R 10.1145/3182644 %7 2018 %D 2018 %J ACM Transactions on Graphics %V 37 %N 2 %& 1 %P 1 - 15 %Z sequence number: 25 %I ACM %C New York, NY %@ false
Thunberg, J., Markdahl, J., Bernard, F., and Goncalves, J. 2018. A Lifting Method for Analyzing Distributed Synchronization on the Unit Sphere. Automatica96.
Export
BibTeX
@article{Thunberg_2018, TITLE = {A Lifting Method for Analyzing Distributed Synchronization on the Unit Sphere}, AUTHOR = {Thunberg, Johan and Markdahl, Johan and Bernard, Florian and Goncalves, Jorge}, LANGUAGE = {eng}, ISSN = {0005-1098}, DOI = {10.1016/j.automatica.2018.07.007}, PUBLISHER = {Elsevier}, ADDRESS = {Amsterdam}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {Automatica}, VOLUME = {96}, PAGES = {253--258}, }
Endnote
%0 Journal Article %A Thunberg, Johan %A Markdahl, Johan %A Bernard, Florian %A Goncalves, Jorge %+ External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations %T A Lifting Method for Analyzing Distributed Synchronization on the Unit Sphere : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5758-A %R 10.1016/j.automatica.2018.07.007 %7 2018 %D 2018 %J Automatica %V 96 %& 253 %P 253 - 258 %I Elsevier %C Amsterdam %@ false
Wolski, K., Giunchi, D., Ye, N., et al. 2018. Dataset and Metrics for Predicting Local Visible Differences. ACM Transactions on Graphics37, 5.
Export
BibTeX
@article{wolski2018dataset, TITLE = {Dataset and Metrics for Predicting Local Visible Differences}, AUTHOR = {Wolski, Krzysztof and Giunchi, Daniele and Ye, Nanyang and Didyk, Piotr and Myszkowski, Karol and Mantiuk, Rados{\textbackslash}l{\textbraceleft}{\textbraceright}aw and Seidel, Hans-Peter and Steed, Anthony and Mantiuk, Rafa{\l} K.}, LANGUAGE = {eng}, ISSN = {0730-0301}, DOI = {10.1145/3196493}, PUBLISHER = {ACM}, ADDRESS = {New York, NY}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {ACM Transactions on Graphics}, VOLUME = {37}, NUMBER = {5}, EID = {172}, }
Endnote
%0 Journal Article %A Wolski, Krzysztof %A Giunchi, Daniele %A Ye, Nanyang %A Didyk, Piotr %A Myszkowski, Karol %A Mantiuk, Rados\l{}aw %A Seidel, Hans-Peter %A Steed, Anthony %A Mantiuk, Rafał K. %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations %T Dataset and Metrics for Predicting Local Visible Differences : %G eng %U http://hdl.handle.net/21.11116/0000-0001-5F75-2 %R 10.1145/3196493 %7 2018 %D 2018 %J ACM Transactions on Graphics %V 37 %N 5 %Z sequence number: 172 %I ACM %C New York, NY %@ false
Xu, W., Chatterjee, A., Zollhöfer, M., et al. 2018a. MonoPerfCap: Human Performance Capture from Monocular Video. ACM Transactions on Graphics37, 2.
Abstract
We present the first marker-less approach for temporally coherent 3D performance capture of a human with general clothing from monocular video. Our approach reconstructs articulated human skeleton motion as well as medium-scale non-rigid surface deformations in general scenes. Human performance capture is a challenging problem due to the large range of articulation, potentially fast motion, and considerable non-rigid deformations, even from multi-view data. Reconstruction from monocular video alone is drastically more challenging, since strong occlusions and the inherent depth ambiguity lead to a highly ill-posed reconstruction problem. We tackle these challenges by a novel approach that employs sparse 2D and 3D human pose detections from a convolutional neural network using a batch-based pose estimation strategy. Joint recovery of per-batch motion allows to resolve the ambiguities of the monocular reconstruction problem based on a low dimensional trajectory subspace. In addition, we propose refinement of the surface geometry based on fully automatically extracted silhouettes to enable medium-scale non-rigid alignment. We demonstrate state-of-the-art performance capture results that enable exciting applications such as video editing and free viewpoint video, previously infeasible from monocular video. Our qualitative and quantitative evaluation demonstrates that our approach significantly outperforms previous monocular methods in terms of accuracy, robustness and scene complexity that can be handled.
Export
BibTeX
@article{Xu_ToG2018, TITLE = {{MonoPerfCap}: Human Performance Capture from Monocular Video}, AUTHOR = {Xu, Weipeng and Chatterjee, Avishek and Zollh{\"o}fer, Michael and Rhodin, Helge and Mehta, Dushyant and Seidel, Hans-Peter and Theobalt, Christian}, LANGUAGE = {eng}, ISSN = {0730-0301}, DOI = {10.1145/3181973}, PUBLISHER = {ACM}, ADDRESS = {New York, NY}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, ABSTRACT = {We present the first marker-less approach for temporally coherent 3D performance capture of a human with general clothing from monocular video. Our approach reconstructs articulated human skeleton motion as well as medium-scale non-rigid surface deformations in general scenes. Human performance capture is a challenging problem due to the large range of articulation, potentially fast motion, and considerable non-rigid deformations, even from multi-view data. Reconstruction from monocular video alone is drastically more challenging, since strong occlusions and the inherent depth ambiguity lead to a highly ill-posed reconstruction problem. We tackle these challenges by a novel approach that employs sparse 2D and 3D human pose detections from a convolutional neural network using a batch-based pose estimation strategy. Joint recovery of per-batch motion allows to resolve the ambiguities of the monocular reconstruction problem based on a low dimensional trajectory subspace. In addition, we propose refinement of the surface geometry based on fully automatically extracted silhouettes to enable medium-scale non-rigid alignment. We demonstrate state-of-the-art performance capture results that enable exciting applications such as video editing and free viewpoint video, previously infeasible from monocular video. Our qualitative and quantitative evaluation demonstrates that our approach significantly outperforms previous monocular methods in terms of accuracy, robustness and scene complexity that can be handled.}, JOURNAL = {ACM Transactions on Graphics}, VOLUME = {37}, NUMBER = {2}, EID = {27}, }
Endnote
%0 Journal Article %A Xu, Weipeng %A Chatterjee, Avishek %A Zollhöfer, Michael %A Rhodin, Helge %A Mehta, Dushyant %A Seidel, Hans-Peter %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T MonoPerfCap: Human Performance Capture from Monocular Video : %G eng %U http://hdl.handle.net/21.11116/0000-0001-E20E-1 %R 10.1145/3181973 %7 2017 %D 2018 %X We present the first marker-less approach for temporally coherent 3D performance capture of a human with general clothing from monocular video. Our approach reconstructs articulated human skeleton motion as well as medium-scale non-rigid surface deformations in general scenes. Human performance capture is a challenging problem due to the large range of articulation, potentially fast motion, and considerable non-rigid deformations, even from multi-view data. Reconstruction from monocular video alone is drastically more challenging, since strong occlusions and the inherent depth ambiguity lead to a highly ill-posed reconstruction problem. We tackle these challenges by a novel approach that employs sparse 2D and 3D human pose detections from a convolutional neural network using a batch-based pose estimation strategy. Joint recovery of per-batch motion allows to resolve the ambiguities of the monocular reconstruction problem based on a low dimensional trajectory subspace. In addition, we propose refinement of the surface geometry based on fully automatically extracted silhouettes to enable medium-scale non-rigid alignment. We demonstrate state-of-the-art performance capture results that enable exciting applications such as video editing and free viewpoint video, previously infeasible from monocular video. Our qualitative and quantitative evaluation demonstrates that our approach significantly outperforms previous monocular methods in terms of accuracy, robustness and scene complexity that can be handled. %K Computer Science, Computer Vision and Pattern Recognition, cs.CV,Computer Science, Graphics, cs.GR %J ACM Transactions on Graphics %V 37 %N 2 %Z sequence number: 27 %I ACM %C New York, NY %@ false
Zayer, R., Mlakar, D., Steinberger, M., and Seidel, H.-P. 2018a. Layered Fields for Natural Tessellations on Surfaces. ACM Transactions on Graphics (Proc. ACM SIGGRAPH Asia 2018)37, 6.
Export
BibTeX
@article{Zayer:2018:LFN, TITLE = {Layered Fields for Natural Tessellations on Surfaces}, AUTHOR = {Zayer, Rhaleb and Mlakar, Daniel and Steinberger, Markus and Seidel, Hans-Peter}, LANGUAGE = {eng}, ISSN = {0730-0301}, ISBN = {978-1-4503-6008-1}, DOI = {10.1145/3272127.3275072}, PUBLISHER = {ACM}, ADDRESS = {New York, NY}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {ACM Transactions on Graphics (Proc. ACM SIGGRAPH Asia)}, VOLUME = {37}, NUMBER = {6}, EID = {264}, BOOKTITLE = {Proceedings of ACM SIGGRAPH Asia 2018}, }
Endnote
%0 Journal Article %A Zayer, Rhaleb %A Mlakar, Daniel %A Steinberger, Markus %A Seidel, Hans-Peter %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T Layered Fields for Natural Tessellations on Surfaces : %G eng %U http://hdl.handle.net/21.11116/0000-0002-E5E0-E %R 10.1145/3272127.3275072 %7 2018 %D 2018 %J ACM Transactions on Graphics %V 37 %N 6 %Z sequence number: 264 %I ACM %C New York, NY %@ false %B Proceedings of ACM SIGGRAPH Asia 2018 %O ACM SIGGRAPH Asia 2018 Tokyo, Japan, December 04 - 07, 2018 SA'18 SA 2018 %@ 978-1-4503-6008-1
Zollhöfer, M., Stotko, P., Görlitz, A., et al. 2018a. State of the Art on 3D Reconstruction with RGB-D Cameras. Computer Graphics Forum (Proc. EUROGRAPHICS 2018)37, 2.
Export
BibTeX
@article{Zollhoefer_EG2018STAR, TITLE = {State of the Art on {3D} Reconstruction with {RGB}-{D} Cameras}, AUTHOR = {Zollh{\"o}fer, Michael and Stotko, Patrick and G{\"o}rlitz, Andreas and Theobalt, Christian and Nie{\ss}ner, Matthias and Klein, Reinhard and Kolb, Andreas}, LANGUAGE = {eng}, ISSN = {0167-7055}, DOI = {10.1111/cgf.13386}, PUBLISHER = {Blackwell-Wiley}, ADDRESS = {Oxford}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {Computer Graphics Forum (Proc. EUROGRAPHICS)}, VOLUME = {37}, NUMBER = {2}, PAGES = {625--652}, BOOKTITLE = {EUROGRAPHICS 2018 STAR -- State of The Art Reports}, EDITOR = {Hildebrandt, Klaus and Theobalt, Christian}, }
Endnote
%0 Journal Article %A Zollhöfer, Michael %A Stotko, Patrick %A Görlitz, Andreas %A Theobalt, Christian %A Nießner, Matthias %A Klein, Reinhard %A Kolb, Andreas %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations External Organizations %T State of the Art on 3D Reconstruction with RGB-D Cameras : %G eng %U http://hdl.handle.net/21.11116/0000-0001-6917-0 %R 10.1111/cgf.13386 %7 2018 %D 2018 %J Computer Graphics Forum %O Computer Graphics Forum : journal of the European Association for Computer Graphics Comput. Graph. Forum %V 37 %N 2 %& 625 %P 625 - 652 %I Blackwell-Wiley %C Oxford %@ false %B EUROGRAPHICS 2018 STAR – State of The Art Reports %O EUROGRAPHICS 2018 EG 2018 The European Association for Computer Graphics 39th Annual Conference ; Delft, The Netherlands, April 16th - 20th, 2018
Zollhöfer, M., Thies, J., Garrido, P., et al. 2018b. State of the Art on Monocular 3D Face Reconstruction, Tracking, and Applications. Computer Graphics Forum (Proc. EUROGRAPHICS 2018)37, 2.
Export
BibTeX
@article{Zollhoefer_EG2018STARb, TITLE = {State of the Art on Monocular {3D} Face Reconstruction, Tracking, and Applications}, AUTHOR = {Zollh{\"o}fer, Michael and Thies, Justus and Garrido, Pablo and Bradley, D. and Beeler, T. and P{\'e}rez, P. and Stamminger, Marc and Nie{\ss}ner, Matthias and Theobalt, Christian}, LANGUAGE = {eng}, ISSN = {0167-7055}, DOI = {10.1111/cgf.13382}, PUBLISHER = {Blackwell-Wiley}, ADDRESS = {Oxford}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, JOURNAL = {Computer Graphics Forum (Proc. EUROGRAPHICS)}, VOLUME = {37}, NUMBER = {2}, PAGES = {523--550}, BOOKTITLE = {EUROGRAPHICS 2018 STAR -- State of The Art Reports}, EDITOR = {Hildebrandt, Klaus and Theobalt, Christian}, }
Endnote
%0 Journal Article %A Zollhöfer, Michael %A Thies, Justus %A Garrido, Pablo %A Bradley, D. %A Beeler, T. %A Pérez, P. %A Stamminger, Marc %A Nießner, Matthias %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations External Organizations External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T State of the Art on Monocular 3D Face Reconstruction, Tracking, and Applications : %G eng %U http://hdl.handle.net/21.11116/0000-0001-691A-D %R 10.1111/cgf.13382 %7 2018 %D 2018 %J Computer Graphics Forum %O Computer Graphics Forum : journal of the European Association for Computer Graphics Comput. Graph. Forum %V 37 %N 2 %& 523 %P 523 - 550 %I Blackwell-Wiley %C Oxford %@ false %B EUROGRAPHICS 2018 STAR – State of The Art Reports %O EUROGRAPHICS 2018 EG 2018 The European Association for Computer Graphics 39th Annual Conference ; Delft, The Netherlands, April 16th - 20th, 2018
Conference Paper
Alldieck, T., Magnor, M.A., Xu, W., Theobalt, C., and Pons-Moll, G. 2018a. Detailed Human Avatars from Monocular Video. 3DV 2018 , International Conference on 3D Vision, IEEE.
Export
BibTeX
@inproceedings{Alldieck_3DV2018, TITLE = {Detailed Human Avatars from Monocular Video}, AUTHOR = {Alldieck, Thiemo and Magnor, Marcus A. and Xu, Weipeng and Theobalt, Christian and Pons-Moll, Gerard}, LANGUAGE = {eng}, ISBN = {978-1-5386-8425-2 ; 978-1-5386-8426-9}, DOI = {10.1109/3DV.2018.00022}, PUBLISHER = {IEEE}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, BOOKTITLE = {3DV 2018 , International Conference on 3D Vision}, PAGES = {98--109}, ADDRESS = {Verona, Italy}, }
Endnote
%0 Conference Proceedings %A Alldieck, Thiemo %A Magnor, Marcus A. %A Xu, Weipeng %A Theobalt, Christian %A Pons-Moll, Gerard %+ Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society %T Detailed Human Avatars from Monocular Video : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5C40-F %R 10.1109/3DV.2018.00022 %D 2018 %B International Conference on 3D Vision %Z date of event: 2018-09-05 - 2018-09-08 %C Verona, Italy %B 3DV 2018 %P 98 - 109 %I IEEE %@ 978-1-5386-8425-2 978-1-5386-8426-9
Alldieck, T., Magnor, M.A., Xu, W., Theobalt, C., and Pons-Moll, G. 2018b. Video Based Reconstruction of 3D People Models. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2018), IEEE.
Export
BibTeX
@inproceedings{alldieck2018video, TITLE = {Video Based Reconstruction of {3D} People Models}, AUTHOR = {Alldieck, Thiemo and Magnor, Marcus A. and Xu, Weipeng and Theobalt, Christian and Pons-Moll, Gerard}, LANGUAGE = {eng}, ISBN = {978-1-5386-6420-9}, DOI = {10.1109/CVPR.2018.00875}, PUBLISHER = {IEEE}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, BOOKTITLE = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2018)}, PAGES = {8387--8397}, ADDRESS = {Salt Lake City, UT, USA}, }
Endnote
%0 Conference Proceedings %A Alldieck, Thiemo %A Magnor, Marcus A. %A Xu, Weipeng %A Theobalt, Christian %A Pons-Moll, Gerard %+ External Organizations External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society %T Video Based Reconstruction of 3D People Models : %G eng %U http://hdl.handle.net/21.11116/0000-0001-1E24-6 %R 10.1109/CVPR.2018.00875 %D 2018 %B 31st IEEE Conference on Computer Vision and Pattern Recognition %Z date of event: 2018-06-18 - 2018-06-22 %C Salt Lake City, UT, USA %B IEEE/CVF Conference on Computer Vision and Pattern Recognition %P 8387 - 8397 %I IEEE %@ 978-1-5386-6420-9
Bernard, F., Theobalt, C., and Moeller, M. 2018a. DS*: Tighter Lifting-Free Convex Relaxations for Quadratic Matching Problems. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2018), IEEE.
Export
BibTeX
@inproceedings{BernardCVPR2018, TITLE = {DS*: {T}ighter Lifting-Free Convex Relaxations for Quadratic Matching Problems}, AUTHOR = {Bernard, Florian and Theobalt, Christian and Moeller, Michael}, LANGUAGE = {eng}, ISBN = {978-1-5386-6420-9}, DOI = {10.1109/CVPR.2018.00453}, PUBLISHER = {IEEE}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, BOOKTITLE = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2018)}, PAGES = {4310--4319}, ADDRESS = {Salt Lake City, UT, USA}, }
Endnote
%0 Conference Proceedings %A Bernard, Florian %A Theobalt, Christian %A Moeller, Michael %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations %T DS*: Tighter Lifting-Free Convex Relaxations for Quadratic Matching Problems : %G eng %U http://hdl.handle.net/21.11116/0000-0002-E92F-4 %R 10.1109/CVPR.2018.00453 %D 2018 %B 31st IEEE Conference on Computer Vision and Pattern Recognition %Z date of event: 2018-06-18 - 2018-06-22 %C Salt Lake City, UT, USA %B IEEE/CVF Conference on Computer Vision and Pattern Recognition %P 4310 - 4319 %I IEEE %@ 978-1-5386-6420-9
Kim, H., Zollhöfer, M., Tewari, A., Thies, J., Richardt, C., and Theobalt, C. 2018b. InverseFaceNet: Deep Monocular Inverse Face Rendering. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2018), IEEE.
Export
BibTeX
@inproceedings{kim2018inverse, TITLE = {{InverseFaceNet}: {D}eep Monocular Inverse Face Rendering}, AUTHOR = {Kim, Hyeongwoo and Zollh{\"o}fer, Michael and Tewari, Ayush and Thies, Justus and Richardt, Christian and Theobalt, Christian}, LANGUAGE = {eng}, ISBN = {978-1-5386-6420-9}, DOI = {10.1109/CVPR.2018.00486}, PUBLISHER = {IEEE}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, BOOKTITLE = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2018)}, PAGES = {4625--4634}, ADDRESS = {Salt Lake City, UT, USA}, }
Endnote
%0 Conference Proceedings %A Kim, Hyeongwoo %A Zollhöfer, Michael %A Tewari, Ayush %A Thies, Justus %A Richardt, Christian %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T InverseFaceNet: Deep Monocular Inverse Face Rendering : %G eng %U http://hdl.handle.net/21.11116/0000-0002-F476-6 %R 10.1109/CVPR.2018.00486 %D 2018 %B 31st IEEE Conference on Computer Vision and Pattern Recognition %Z date of event: 2018-06-18 - 2018-06-22 %C Salt Lake City, UT, USA %B IEEE/CVF Conference on Computer Vision and Pattern Recognition %P 4625 - 4634 %I IEEE %@ 978-1-5386-6420-9
Mehta, D., Sotnychenko, O., Mueller, F., et al. 2018a. Single-Shot Multi-person 3D Pose Estimation from Monocular RGB. 3DV 2018 , International Conference on 3D Vision, IEEE.
Export
BibTeX
@inproceedings{Mehta_3DV2018, TITLE = {Single-Shot Multi-person {3D} Pose Estimation from Monocular {RGB}}, AUTHOR = {Mehta, Dushyant and Sotnychenko, Oleksandr and Mueller, Franziska and Xu, Weipeng and Sridhar, Srinath and Pons-Moll, Gerard and Theobalt, Christian}, LANGUAGE = {eng}, ISBN = {978-1-5386-8425-2 ; 978-1-5386-8426-9}, DOI = {10.1109/3DV.2018.00024}, PUBLISHER = {IEEE}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, BOOKTITLE = {3DV 2018 , International Conference on 3D Vision}, PAGES = {120--130}, ADDRESS = {Verona, Italy}, }
Endnote
%0 Conference Proceedings %A Mehta, Dushyant %A Sotnychenko, Oleksandr %A Mueller, Franziska %A Xu, Weipeng %A Sridhar, Srinath %A Pons-Moll, Gerard %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T Single-Shot Multi-person 3D Pose Estimation from Monocular RGB : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5C46-9 %R 10.1109/3DV.2018.00024 %D 2018 %B International Conference on 3D Vision %Z date of event: 2018-09-05 - 2018-09-08 %C Verona, Italy %B 3DV 2018 %P 120 - 130 %I IEEE %@ 978-1-5386-8425-2 978-1-5386-8426-9
Mehta, D., Sotnychenko, O., Mueller, F., et al. 2018b. Demo of XNect: Real-time Multi-person 3D Human Pose Estimation with a Single RGB Camera. ECCV 2018 Demo Sessions.
Export
BibTeX
@inproceedings{XNectDemo_ECCV2018, TITLE = {Demo of {XNect}: Real-time Multi-person {3D} Human Pose Estimation with a Single {RGB} Camera}, AUTHOR = {Mehta, Dushyant and Sotnychenko, Oleksandr and Mueller, Franziska and Rhodin, Helge and Xu, Weipeng and Pons-Moll, Gerard and Theobalt, Christian}, LANGUAGE = {eng}, URL = {http://gvv.mpi-inf.mpg.de/projects/XNectDemo/}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, BOOKTITLE = {ECCV 2018 Demo Sessions}, ADDRESS = {Munich, Germany}, }
Endnote
%0 Conference Proceedings %A Mehta, Dushyant %A Sotnychenko, Oleksandr %A Mueller, Franziska %A Rhodin, Helge %A Xu, Weipeng %A Pons-Moll, Gerard %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T Demo of XNect: Real-time Multi-person 3D Human Pose Estimation with a Single RGB Camera : %G eng %U http://hdl.handle.net/21.11116/0000-0002-F4DC-3 %U http://gvv.mpi-inf.mpg.de/projects/XNectDemo/ %D 2018 %B European Conference on Computer Vision %Z date of event: 2018-09-08 - 2018-09-14 %C Munich, Germany %B ECCV 2018 Demo Sessions %U http://gvv.mpi-inf.mpg.de/projects/XNectDemo/
Meka, A., Maximov, M., Zollhöfer, M., et al. 2018a. LIME: Live Intrinsic Material Estimation. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2018), IEEE.
Export
BibTeX
@inproceedings{Meka:2018, TITLE = {{LIME}: {L}ive Intrinsic Material Estimation}, AUTHOR = {Meka, Abhimitra and Maximov, Maxim and Zollh{\"o}fer, Michael and Chatterjee, Avishek and Seidel, Hans-Peter and Richardt, Christian and Theobalt, Christian}, LANGUAGE = {eng}, ISBN = {978-1-5386-6420-9}, DOI = {10.1109/CVPR.2018.00661}, PUBLISHER = {IEEE}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, BOOKTITLE = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2018)}, PAGES = {6315--6324}, ADDRESS = {Salt Lake City, UT, USA}, }
Endnote
%0 Conference Proceedings %A Meka, Abhimitra %A Maximov, Maxim %A Zollhöfer, Michael %A Chatterjee, Avishek %A Seidel, Hans-Peter %A Richardt, Christian %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T LIME: Live Intrinsic Material Estimation : %G eng %U http://hdl.handle.net/21.11116/0000-0002-F391-7 %R 10.1109/CVPR.2018.00661 %D 2018 %B 31st IEEE Conference on Computer Vision and Pattern Recognition %Z date of event: 2018-06-18 - 2018-06-22 %C Salt Lake City, UT, USA %B IEEE/CVF Conference on Computer Vision and Pattern Recognition %P 6315 - 6324 %I IEEE %@ 978-1-5386-6420-9 %U http://gvv.mpi-inf.mpg.de/projects/LIME/
Mueller, F., Bernard, F., Sotnychenko, O., et al. 2018. GANerated Hands for Real-Time 3D Hand Tracking from Monocular RGB. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2018), IEEE.
Export
BibTeX
@inproceedings{Mueller_CVPR2018, TITLE = {{GANerated} Hands for Real-Time {3D} Hand Tracking from Monocular {RGB}}, AUTHOR = {Mueller, Franziska and Bernard, Florian and Sotnychenko, Oleksandr and Mehta, Dushyant and Sridhar, Srinath and Casas, Dan and Theobalt, Christian}, LANGUAGE = {eng}, ISBN = {978-1-5386-6420-9}, DOI = {10.1109/CVPR.2018.00013}, PUBLISHER = {IEEE}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, BOOKTITLE = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2018)}, PAGES = {49--59}, ADDRESS = {Salt Lake City, UT, USA}, }
Endnote
%0 Conference Proceedings %A Mueller, Franziska %A Bernard, Florian %A Sotnychenko, Oleksandr %A Mehta, Dushyant %A Sridhar, Srinath %A Casas, Dan %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T GANerated Hands for Real-Time 3D Hand Tracking from Monocular RGB : %G eng %U http://hdl.handle.net/21.11116/0000-0002-EFFA-8 %R 10.1109/CVPR.2018.00013 %D 2018 %B 31st IEEE Conference on Computer Vision and Pattern Recognition %Z date of event: 2018-06-18 - 2018-06-22 %C Salt Lake City, UT, USA %B IEEE/CVF Conference on Computer Vision and Pattern Recognition %P 49 - 59 %I IEEE %@ 978-1-5386-6420-9
Myszkowski, K., Tursun, O.T., Kellnhofer, P., et al. 2018. Perceptual Display: Apparent Enhancement of Scene Detail and Depth. Electronic Imaging (Proc. HVEI 2018), SPIE/IS&T.
(Keynote Talk)
Abstract
Predicting human visual perception of image differences has several applications such as compression, rendering, editing and retargeting. Current approaches however, ignore the fact that the human visual system compensates for geometric transformations, e.g. we see that an image and a rotated copy are identical. Instead, they will report a large, false-positive difference. At the same time, if the transformations become too strong or too spatially incoherent, comparing two images indeed gets increasingly difficult. Between these two extremes, we propose a system to quantify the effect of transformations, not only on the perception of image differences, but also on saliency and motion parallax. To this end, we first fit local homographies to a given optical flow field and then convert this field into a field of elementary transformations such as translation, rotation, scaling, and perspective. We conduct a perceptual experiment quantifying the increase of difficulty when compensating for elementary transformations. Transformation entropy is proposed as a novel measure of complexity in a flow field. This representation is then used for applications, such as comparison of non-aligned images, where transformations cause threshold elevation, and detection of salient transformations.
Export
BibTeX
@inproceedings{Myszkowski2018Perceptual, TITLE = {Perceptual Display: Apparent Enhancement of Scene Detail and Depth}, AUTHOR = {Myszkowski, Karol and Tursun, Okan Tarhan and Kellnhofer, Petr and Templin, Krzysztof and Arabadzhiyska, Elena and Didyk, Piotr and Seidel, Hans-Peter}, LANGUAGE = {eng}, ISSN = {2470-1173}, DOI = {10.2352/ISSN.2470-1173.2018.14.HVEI-501}, PUBLISHER = {SPIE/IS\&T}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {Predicting human visual perception of image differences has several applications such as compression, rendering, editing and retargeting. Current approaches however, ignore the fact that the human visual system compensates for geometric transformations, e.g. we see that an image and a rotated copy are identical. Instead, they will report a large, false-positive difference. At the same time, if the transformations become too strong or too spatially incoherent, comparing two images indeed gets increasingly difficult. Between these two extremes, we propose a system to quantify the effect of transformations, not only on the perception of image differences, but also on saliency and motion parallax. To this end, we first fit local homographies to a given optical flow field and then convert this field into a field of elementary transformations such as translation, rotation, scaling, and perspective. We conduct a perceptual experiment quantifying the increase of difficulty when compensating for elementary transformations. Transformation entropy is proposed as a novel measure of complexity in a flow field. This representation is then used for applications, such as comparison of non-aligned images, where transformations cause threshold elevation, and detection of salient transformations.}, BOOKTITLE = {Human Vision and Electronic Imaging (HVEI 2018)}, PAGES = {1--10}, EID = {501}, JOURNAL = {Electronic Imaging (Proc. HVEI)}, VOLUME = {2018}, ADDRESS = {San Francisco, CA, USA}, }
Endnote
%0 Conference Proceedings %A Myszkowski, Karol %A Tursun, Okan Tarhan %A Kellnhofer, Petr %A Templin, Krzysztof %A Arabadzhiyska, Elena %A Didyk, Piotr %A Seidel, Hans-Peter %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T Perceptual Display: Apparent Enhancement of Scene Detail and Depth : %G eng %U http://hdl.handle.net/21.11116/0000-0001-5F64-5 %R 10.2352/ISSN.2470-1173.2018.14.HVEI-501 %D 2018 %B Human Vision and Electronic Imaging %Z date of event: 2018-01-28 - 2018-02-02 %C San Francisco, CA, USA %X Predicting human visual perception of image differences has several applications such as compression, rendering, editing and retargeting. Current approaches however, ignore the fact that the human visual system compensates for geometric transformations, e.g. we see that an image and a rotated copy are identical. Instead, they will report a large, false-positive difference. At the same time, if the transformations become too strong or too spatially incoherent, comparing two images indeed gets increasingly difficult. Between these two extremes, we propose a system to quantify the effect of transformations, not only on the perception of image differences, but also on saliency and motion parallax. To this end, we first fit local homographies to a given optical flow field and then convert this field into a field of elementary transformations such as translation, rotation, scaling, and perspective. We conduct a perceptual experiment quantifying the increase of difficulty when compensating for elementary transformations. Transformation entropy is proposed as a novel measure of complexity in a flow field. This representation is then used for applications, such as comparison of non-aligned images, where transformations cause threshold elevation, and detection of salient transformations. %B Human Vision and Electronic Imaging %P 1 - 10 %Z sequence number: 501 %I SPIE/IS&T %J Electronic Imaging %V 2018 %@ false
Öztireli, A.C. and Singh, G. 2018. Sampling Analysis Using Correlations for Monte Carlo Rendering. SIGGRAPH Asia 2018 Courses (ACM SIGGRAPH Asia 2018), ACM.
Export
BibTeX
@inproceedings{Oeztireli_SIGGRAPHASIA18, TITLE = {Sampling Analysis Using Correlations for {M}onte {C}arlo Rendering}, AUTHOR = {{\"O}ztireli, A. Cengiz and Singh, Gurprit}, LANGUAGE = {eng}, DOI = {10.1145/3277644.3277783}, PUBLISHER = {ACM}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, BOOKTITLE = {SIGGRAPH Asia 2018 Courses (ACM SIGGRAPH Asia 2018)}, PAGES = {1--48}, EID = {16}, ADDRESS = {Tokyo, Japan}, }
Endnote
%0 Conference Proceedings %A Öztireli, A. Cengiz %A Singh, Gurprit %+ External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T Sampling Analysis Using Correlations for Monte Carlo Rendering : %G eng %U http://hdl.handle.net/21.11116/0000-0002-9F4D-6 %R 10.1145/3277644.3277783 %D 2018 %B 11th ACM SIGGRAPH Conference and Exhibition on Computer Graphics and Interactive Techniques in Asia %Z date of event: 2018-12-04 - 2018-12-07 %C Tokyo, Japan %B SIGGRAPH Asia 2018 Courses %P 1 - 48 %Z sequence number: 16 %I ACM
Robertini, N., Bernard, F., Xu, W., and Theobalt, C. 2018. Illumination-Invariant Robust Multiview 3D Human Motion Capture. 2018 IEEE Winter Conference on Applications of Computer Vision (WACV 2018), IEEE.
Export
BibTeX
@inproceedings{Robertini_WACV2018, TITLE = {Illumination-Invariant Robust Multiview {3D} Human Motion Capture}, AUTHOR = {Robertini, Nadia and Bernard, Florian and Xu, Weipeng and Theobalt, Christian}, LANGUAGE = {eng}, ISBN = {978-1-5386-4886-5}, DOI = {10.1109/WACV.2018.00185}, PUBLISHER = {IEEE}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, BOOKTITLE = {2018 IEEE Winter Conference on Applications of Computer Vision (WACV 2018)}, PAGES = {1661--1670}, ADDRESS = {Lake Tahoe, NV, USA}, }
Endnote
%0 Conference Proceedings %A Robertini, Nadia %A Bernard, Florian %A Xu, Weipeng %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T Illumination-Invariant Robust Multiview 3D Human Motion Capture : %G eng %U http://hdl.handle.net/21.11116/0000-0001-A474-3 %R 10.1109/WACV.2018.00185 %D 2018 %B IEEE Winter Conference on Applications of Computer Vision %Z date of event: 2018-03-12 - 2018-03-15 %C Lake Tahoe, NV, USA %B 2018 IEEE Winter Conference on Applications of Computer Vision %P 1661 - 1670 %I IEEE %@ 978-1-5386-4886-5
Sarkar, K., Bernard, F., Varanasi, K., Theobalt, C., and Stricker, D. 2018. Structured Low-Rank Matrix Factorization for Point-Cloud Denoising. 3DV 2018 , International Conference on 3D Vision, IEEE.
Export
BibTeX
@inproceedings{Sarkar_3DV2018, TITLE = {Structured Low-Rank Matrix Factorization for Point-Cloud Denoising}, AUTHOR = {Sarkar, Kripasindhu and Bernard, Florian and Varanasi, Kiran and Theobalt, Christian and Stricker, Didier}, LANGUAGE = {eng}, ISBN = {978-1-5386-8425-2 ; 978-1-5386-8426-9}, DOI = {10.1109/3DV.2018.00058}, PUBLISHER = {IEEE}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, BOOKTITLE = {3DV 2018 , International Conference on 3D Vision}, PAGES = {444--453}, ADDRESS = {Verona, Italy}, }
Endnote
%0 Conference Proceedings %A Sarkar, Kripasindhu %A Bernard, Florian %A Varanasi, Kiran %A Theobalt, Christian %A Stricker, Didier %+ External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations %T Structured Low-Rank Matrix Factorization for Point-Cloud Denoising : %G eng %U http://hdl.handle.net/21.11116/0000-0002-D62C-C %R 10.1109/3DV.2018.00058 %D 2018 %B International Conference on 3D Vision %Z date of event: 2018-09-05 - 2018-09-08 %C Verona, Italy %B 3DV 2018 %P 444 - 453 %I IEEE %@ 978-1-5386-8425-2 978-1-5386-8426-9
Shekhar, S., Kunz Beigpour, S., Ziegler, M., et al. 2018. Light-Field Intrinsic Dataset. Proceedings of the British Machine Vision Conference 2018 (BMVC), British Machine Vision Association.
Export
BibTeX
@inproceedings{Shekhar_BMVC2018, TITLE = {Light-Field Intrinsic Dataset}, AUTHOR = {Shekhar, Sumit and Kunz Beigpour, Shida and Ziegler, Matthias and Chwesiuk, Micha{\l} and Pale{\'n}, Dawid and Myszkowski, Karol and Keinert, Joachim and Mantiuk, Rados{\l}aw and Didyk, Piotr}, LANGUAGE = {eng}, URL = {http://bmvc2018.org/programme/BMVC2018.zip}, PUBLISHER = {British Machine Vision Association}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, BOOKTITLE = {Proceedings of the British Machine Vision Conference 2018 (BMVC)}, EID = {0120}, ADDRESS = {Newcastle, UK}, }
Endnote
%0 Conference Proceedings %A Shekhar, Sumit %A Kunz Beigpour, Shida %A Ziegler, Matthias %A Chwesiuk, Michał %A Paleń, Dawid %A Myszkowski, Karol %A Keinert, Joachim %A Mantiuk, Radosław %A Didyk, Piotr %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T Light-Field Intrinsic Dataset : %G eng %U http://hdl.handle.net/21.11116/0000-0002-0E38-1 %D 2018 %B British Machine Vision Conference 2018 (BMVC) %Z date of event: 2018-09-03 - 2018-09-06 %C Newcastle, UK %K Forschungsgruppe Geiger %B Proceedings of the British Machine Vision Conference 2018 (BMVC) %Z sequence number: 0120 %I British Machine Vision Association %U http://bmvc2018.org/programme/BMVC2018.zip
Soliman, M., Mueller, F., Hegemann, L., Roo, J.S., Theobalt, C., and Steimle, J. 2018. FingerInput: Capturing Expressive Single-Hand Thumb-to-Finger Microgestures. ISS’18, 13th ACM International Conference on Interactive Surfaces and Spaces, ACM.
Export
BibTeX
@inproceedings{Soliman_ISS2018, TITLE = {{FingerInput}: {C}apturing Expressive Single-Hand Thumb-to-Finger Microgestures}, AUTHOR = {Soliman, Mohamed and Mueller, Franziska and Hegemann, Lena and Roo, Joan Sol and Theobalt, Christian and Steimle, J{\"u}rgen}, LANGUAGE = {eng}, ISBN = {978-1-4503-5694-7}, DOI = {10.1145/3279778.3279799}, PUBLISHER = {ACM}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, BOOKTITLE = {ISS'18, 13th ACM International Conference on Interactive Surfaces and Spaces}, PAGES = {177--187}, ADDRESS = {Tokyo, Japan}, }
Endnote
%0 Conference Proceedings %A Soliman, Mohamed %A Mueller, Franziska %A Hegemann, Lena %A Roo, Joan Sol %A Theobalt, Christian %A Steimle, Jürgen %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T FingerInput: Capturing Expressive Single-Hand Thumb-to-Finger Microgestures : %G eng %U http://hdl.handle.net/21.11116/0000-0002-A36D-C %R 10.1145/3279778.3279799 %D 2018 %B 13th ACM International Conference on Interactive Surfaces and Spaces %Z date of event: 2018-11-25 - 2018-11-28 %C Tokyo, Japan %B ISS'18 %P 177 - 187 %I ACM %@ 978-1-4503-5694-7
Sun, Q., Tewari, A., Xu, W., Fritz, M., Theobalt, C., and Schiele, B. 2018a. A Hybrid Model for Identity Obfuscation by Face Replacement. Computer Vision -- ECCV 2018, Springer.
Export
BibTeX
@inproceedings{Sun_ECCV2018, TITLE = {A Hybrid Model for Identity Obfuscation by Face Replacement}, AUTHOR = {Sun, Qianru and Tewari, Ayush and Xu, Weipeng and Fritz, Mario and Theobalt, Christian and Schiele, Bernt}, LANGUAGE = {eng}, ISBN = {978-3-030-01245-8}, DOI = {10.1007/978-3-030-01246-5_34}, PUBLISHER = {Springer}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, BOOKTITLE = {Computer Vision -- ECCV 2018}, PAGES = {570--586}, SERIES = {Lecture Notes in Computer Science}, VOLUME = {11205}, ADDRESS = {Munich, Germany}, }
Endnote
%0 Conference Proceedings %A Sun, Qianru %A Tewari, Ayush %A Xu, Weipeng %A Fritz, Mario %A Theobalt, Christian %A Schiele, Bernt %+ Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society %T A Hybrid Model for Identity Obfuscation by Face Replacement : %G eng %U http://hdl.handle.net/21.11116/0000-0001-3F9B-B %R 10.1007/978-3-030-01246-5_34 %D 2018 %B 15th European Conference on Computer Vision %Z date of event: 2018-09-08 - 2018-09-14 %C Munich, Germany %B Computer Vision -- ECCV 2018 %P 570 - 586 %I Springer %@ 978-3-030-01245-8 %B Lecture Notes in Computer Science %N 11205
Tewari, A., Zollhöfer, M., Garrido, P., et al. 2018b. Self-Supervised Multi-level Face Model Learning for Monocular Reconstruction at Over 250 Hz. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2018), IEEE.
Export
BibTeX
@inproceedings{Teware_CVPR2018, TITLE = {Self-Supervised Multi-level Face Model Learning for Monocular Reconstruction at Over 250 {Hz}}, AUTHOR = {Tewari, Ayush and Zollh{\"o}fer, Michael and Garrido, Pablo and Bernard, Florian and Kim, Hyeongwoo and P{\'e}rez, Patrick and Theobalt, Christian}, LANGUAGE = {eng}, ISBN = {978-1-5386-6420-9}, DOI = {10.1109/CVPR.2018.00270}, PUBLISHER = {IEEE}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, BOOKTITLE = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2018)}, PAGES = {2549--2559}, ADDRESS = {Salt Lake City, UT, USA}, }
Endnote
%0 Conference Proceedings %A Tewari, Ayush %A Zollhöfer, Michael %A Garrido, Pablo %A Bernard, Florian %A Kim, Hyeongwoo %A Pérez, Patrick %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T Self-Supervised Multi-level Face Model Learning for Monocular Reconstruction at Over 250 Hz : %G eng %U http://hdl.handle.net/21.11116/0000-0002-EF56-1 %R 10.1109/CVPR.2018.00270 %D 2018 %B 31st IEEE Conference on Computer Vision and Pattern Recognition %Z date of event: 2018-06-18 - 2018-06-22 %C Salt Lake City, UT, USA %B IEEE/CVF Conference on Computer Vision and Pattern Recognition %P 2549 - 2559 %I IEEE %@ 978-1-5386-6420-9
Tretschk, E., Oh, S.J., and Fritz, M. 2018. Sequential Attacks on Agents for Long-Term Adversarial Goals. 2. ACM Computer Science in Cars Symposium (CSCS 2018).
Export
BibTeX
@inproceedings{Tretschk_CSCS2018, TITLE = {Sequential Attacks on Agents for Long-Term Adversarial Goals}, AUTHOR = {Tretschk, Edgar and Oh, Seong Joon and Fritz, Mario}, LANGUAGE = {eng}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, BOOKTITLE = {2. ACM Computer Science in Cars Symposium (CSCS 2018)}, ADDRESS = {Munich, Germany}, }
Endnote
%0 Conference Proceedings %A Tretschk, Edgar %A Oh, Seong Joon %A Fritz, Mario %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society %T Sequential Attacks on Agents for Long-Term Adversarial Goals : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5B46-A %D 2018 %B 2. ACM Computer Science in Cars Symposium %Z date of event: 2018-09-13 - 2018-09-14 %C Munich, Germany %B 2. ACM Computer Science in Cars Symposium
Winter, M., Mlakar, D., Zayer, R., Seidel, H.-P., and Steinberger, M. 2018. faimGraph: High Performance Management of Fully-Dynamic Graphs Under Tight Memory Constraints on the GPU. The International Conference for High Performance Computing, Networking, Storage, and Analysis (SC 2018), IEEE.
Export
BibTeX
@inproceedings{Winter:2018:FHP, TITLE = {{faimGraph}: {H}igh Performance Management of Fully-Dynamic Graphs Under Tight Memory Constraints on the {GPU}}, AUTHOR = {Winter, Martin and Mlakar, Daniel and Zayer, Rhaleb and Seidel, Hans-Peter and Steinberger, Markus}, LANGUAGE = {eng}, ISBN = {978-1-5386-8384-2}, URL = {http://conferences.computer.org/sc/2018/#!/home}, PUBLISHER = {IEEE}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, BOOKTITLE = {The International Conference for High Performance Computing, Networking, Storage, and Analysis (SC 2018)}, PAGES = {754--766}, ADDRESS = {Dallas, TX, USA}, }
Endnote
%0 Conference Proceedings %A Winter, Martin %A Mlakar, Daniel %A Zayer, Rhaleb %A Seidel, Hans-Peter %A Steinberger, Markus %+ External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations %T faimGraph: High Performance Management of Fully-Dynamic Graphs Under Tight Memory Constraints on the GPU : %G eng %U http://hdl.handle.net/21.11116/0000-0002-E5E6-8 %D 2018 %B The International Conference for High Performance Computing, Networking, Storage, and Analysis %Z date of event: 2018-11-11 - 2018-11-16 %C Dallas, TX, USA %B The International Conference for High Performance Computing, Networking, Storage, and Analysis %P 754 - 766 %I IEEE %@ 978-1-5386-8384-2
Paper
Alldieck, T., Magnor, M.A., Xu, W., Theobalt, C., and Pons-Moll, G. 2018c. Detailed Human Avatars from Monocular Video. http://arxiv.org/abs/1808.01338.
(arXiv: 1808.01338)
Abstract
We present a novel method for high detail-preserving human avatar creation from monocular video. A parameterized body model is refined and optimized to maximally resemble subjects from a video showing them from all sides. Our avatars feature a natural face, hairstyle, clothes with garment wrinkles, and high-resolution texture. Our paper contributes facial landmark and shading-based human body shape refinement, a semantic texture prior, and a novel texture stitching strategy, resulting in the most sophisticated-looking human avatars obtained from a single video to date. Numerous results show the robustness and versatility of our method. A user study illustrates its superiority over the state-of-the-art in terms of identity preservation, level of detail, realism, and overall user preference.
Export
BibTeX
@online{Alldieck_arXiv1808.01338, TITLE = {Detailed Human Avatars from Monocular Video}, AUTHOR = {Alldieck, Thiemo and Magnor, Marcus A. and Xu, Weipeng and Theobalt, Christian and Pons-Moll, Gerard}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1808.01338}, EPRINT = {1808.01338}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {We present a novel method for high detail-preserving human avatar creation from monocular video. A parameterized body model is refined and optimized to maximally resemble subjects from a video showing them from all sides. Our avatars feature a natural face, hairstyle, clothes with garment wrinkles, and high-resolution texture. Our paper contributes facial landmark and shading-based human body shape refinement, a semantic texture prior, and a novel texture stitching strategy, resulting in the most sophisticated-looking human avatars obtained from a single video to date. Numerous results show the robustness and versatility of our method. A user study illustrates its superiority over the state-of-the-art in terms of identity preservation, level of detail, realism, and overall user preference.}, }
Endnote
%0 Report %A Alldieck, Thiemo %A Magnor, Marcus A. %A Xu, Weipeng %A Theobalt, Christian %A Pons-Moll, Gerard %+ Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society %T Detailed Human Avatars from Monocular Video : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5E0F-6 %U http://arxiv.org/abs/1808.01338 %D 2018 %X We present a novel method for high detail-preserving human avatar creation from monocular video. A parameterized body model is refined and optimized to maximally resemble subjects from a video showing them from all sides. Our avatars feature a natural face, hairstyle, clothes with garment wrinkles, and high-resolution texture. Our paper contributes facial landmark and shading-based human body shape refinement, a semantic texture prior, and a novel texture stitching strategy, resulting in the most sophisticated-looking human avatars obtained from a single video to date. Numerous results show the robustness and versatility of our method. A user study illustrates its superiority over the state-of-the-art in terms of identity preservation, level of detail, realism, and overall user preference. %K Computer Science, Computer Vision and Pattern Recognition, cs.CV
Alldieck, T., Magnor, M.A., Xu, W., Theobalt, C., and Pons-Moll, G. 2018d. Video Based Reconstruction of 3D People Models. http://arxiv.org/abs/1803.04758.
(arXiv: 1803.04758)
Abstract
This paper describes how to obtain accurate 3D body models and texture of arbitrary people from a single, monocular video in which a person is moving. Based on a parametric body model, we present a robust processing pipeline achieving 3D model fits with 5mm accuracy also for clothed people. Our main contribution is a method to nonrigidly deform the silhouette cones corresponding to the dynamic human silhouettes, resulting in a visual hull in a common reference frame that enables surface reconstruction. This enables efficient estimation of a consensus 3D shape, texture and implanted animation skeleton based on a large number of frames. We present evaluation results for a number of test subjects and analyze overall performance. Requiring only a smartphone or webcam, our method enables everyone to create their own fully animatable digital double, e.g., for social VR applications or virtual try-on for online fashion shopping.
Export
BibTeX
@online{Alldieck_arXiv1803.04758, TITLE = {Video Based Reconstruction of {3D} People Models}, AUTHOR = {Alldieck, Thiemo and Magnor, Marcus A. and Xu, Weipeng and Theobalt, Christian and Pons-Moll, Gerard}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1803.04758}, EPRINT = {1803.04758}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {This paper describes how to obtain accurate 3D body models and texture of arbitrary people from a single, monocular video in which a person is moving. Based on a parametric body model, we present a robust processing pipeline achieving 3D model fits with 5mm accuracy also for clothed people. Our main contribution is a method to nonrigidly deform the silhouette cones corresponding to the dynamic human silhouettes, resulting in a visual hull in a common reference frame that enables surface reconstruction. This enables efficient estimation of a consensus 3D shape, texture and implanted animation skeleton based on a large number of frames. We present evaluation results for a number of test subjects and analyze overall performance. Requiring only a smartphone or webcam, our method enables everyone to create their own fully animatable digital double, e.g., for social VR applications or virtual try-on for online fashion shopping.}, }
Endnote
%0 Report %A Alldieck, Thiemo %A Magnor, Marcus A. %A Xu, Weipeng %A Theobalt, Christian %A Pons-Moll, Gerard %+ External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society D2 Extern %T Video Based Reconstruction of 3D People Models : %G eng %U http://hdl.handle.net/21.11116/0000-0001-40CD-0 %U http://arxiv.org/abs/1803.04758 %D 2018 %X This paper describes how to obtain accurate 3D body models and texture of arbitrary people from a single, monocular video in which a person is moving. Based on a parametric body model, we present a robust processing pipeline achieving 3D model fits with 5mm accuracy also for clothed people. Our main contribution is a method to nonrigidly deform the silhouette cones corresponding to the dynamic human silhouettes, resulting in a visual hull in a common reference frame that enables surface reconstruction. This enables efficient estimation of a consensus 3D shape, texture and implanted animation skeleton based on a large number of frames. We present evaluation results for a number of test subjects and analyze overall performance. Requiring only a smartphone or webcam, our method enables everyone to create their own fully animatable digital double, e.g., for social VR applications or virtual try-on for online fashion shopping. %K Computer Science, Computer Vision and Pattern Recognition, cs.CV
Bernard, F., Thunberg, J., Goncalves, J., and Theobalt, C. 2018b. Synchronisation of Partial Multi-Matchings via Non-negative Factorisations. http://arxiv.org/abs/1803.06320.
(arXiv: 1803.06320)
Abstract
In this work we study permutation synchronisation for the challenging case of partial permutations, which plays an important role for the problem of matching multiple objects (e.g. images or shapes). The term synchronisation refers to the property that the set of pairwise matchings is cycle-consistent, i.e. in the full matching case all compositions of pairwise matchings over cycles must be equal to the identity. Motivated by clustering and matrix factorisation perspectives of cycle-consistency, we derive an algorithm to tackle the permutation synchronisation problem based on non-negative factorisations. In order to deal with the inherent non-convexity of the permutation synchronisation problem, we use an initialisation procedure based on a novel rotation scheme applied to the solution of the spectral relaxation. Moreover, this rotation scheme facilitates a convenient Euclidean projection to obtain a binary solution after solving our relaxed problem. In contrast to state-of-the-art methods, our approach is guaranteed to produce cycle-consistent results. We experimentally demonstrate the efficacy of our method and show that it achieves better results compared to existing methods.
Export
BibTeX
@online{Bernard_arXiv1803.06320, TITLE = {Synchronisation of Partial Multi-Matchings via Non-negative Factorisations}, AUTHOR = {Bernard, Florian and Thunberg, Johan and Goncalves, Jorge and Theobalt, Christian}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1803.06320}, EPRINT = {1803.06320}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {In this work we study permutation synchronisation for the challenging case of partial permutations, which plays an important role for the problem of matching multiple objects (e.g. images or shapes). The term synchronisation refers to the property that the set of pairwise matchings is cycle-consistent, i.e. in the full matching case all compositions of pairwise matchings over cycles must be equal to the identity. Motivated by clustering and matrix factorisation perspectives of cycle-consistency, we derive an algorithm to tackle the permutation synchronisation problem based on non-negative factorisations. In order to deal with the inherent non-convexity of the permutation synchronisation problem, we use an initialisation procedure based on a novel rotation scheme applied to the solution of the spectral relaxation. Moreover, this rotation scheme facilitates a convenient Euclidean projection to obtain a binary solution after solving our relaxed problem. In contrast to state-of-the-art methods, our approach is guaranteed to produce cycle-consistent results. We experimentally demonstrate the efficacy of our method and show that it achieves better results compared to existing methods.}, }
Endnote
%0 Report %A Bernard, Florian %A Thunberg, Johan %A Goncalves, Jorge %A Theobalt, Christian %+ External Organizations External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T Synchronisation of Partial Multi-Matchings via Non-negative Factorisations : %G eng %U http://hdl.handle.net/21.11116/0000-0001-40BC-3 %U http://arxiv.org/abs/1803.06320 %D 2018 %X In this work we study permutation synchronisation for the challenging case of partial permutations, which plays an important role for the problem of matching multiple objects (e.g. images or shapes). The term synchronisation refers to the property that the set of pairwise matchings is cycle-consistent, i.e. in the full matching case all compositions of pairwise matchings over cycles must be equal to the identity. Motivated by clustering and matrix factorisation perspectives of cycle-consistency, we derive an algorithm to tackle the permutation synchronisation problem based on non-negative factorisations. In order to deal with the inherent non-convexity of the permutation synchronisation problem, we use an initialisation procedure based on a novel rotation scheme applied to the solution of the spectral relaxation. Moreover, this rotation scheme facilitates a convenient Euclidean projection to obtain a binary solution after solving our relaxed problem. In contrast to state-of-the-art methods, our approach is guaranteed to produce cycle-consistent results. We experimentally demonstrate the efficacy of our method and show that it achieves better results compared to existing methods. %K Computer Science, Computer Vision and Pattern Recognition, cs.CV,Mathematics, Optimization and Control, math.OC,Statistics, Machine Learning, stat.ML
Bernard, F., Thunberg, J., Swoboda, P., and Theobalt, C. 2018c. Higher-order Projected Power Iterations for Scalable Multi-Matching. http://arxiv.org/abs/1811.10541.
(arXiv: 1811.10541)
Abstract
The matching of multiple objects (e.g. shapes or images) is a fundamental problem in vision and graphics. In order to robustly handle ambiguities, noise and repetitive patterns in challenging real-world settings, it is essential to take geometric consistency between points into account. Computationally, the multi-matching problem is difficult. It can be phrased as simultaneously solving multiple (NP-hard) quadratic assignment problems (QAPs) that are coupled via cycle-consistency constraints. The main limitations of existing multi-matching methods are that they either ignore geometric consistency and thus have limited robustness, or they are restricted to small-scale problems due to their (relatively) high computational cost. We address these shortcomings by introducing a Higher-order Projected Power Iteration method, which is (i) efficient and scales to tens of thousands of points, (ii) straightforward to implement, (iii) able to incorporate geometric consistency, and (iv) guarantees cycle-consistent multi-matchings. Experimentally we show that our approach is superior to existing methods.
Export
BibTeX
@online{Bernard_arXIv1811.10541, TITLE = {Higher-order Projected Power Iterations for Scalable Multi-Matching}, AUTHOR = {Bernard, Florian and Thunberg, Johan and Swoboda, Paul and Theobalt, Christian}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1811.10541}, EPRINT = {1811.10541}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {The matching of multiple objects (e.g. shapes or images) is a fundamental problem in vision and graphics. In order to robustly handle ambiguities, noise and repetitive patterns in challenging real-world settings, it is essential to take geometric consistency between points into account. Computationally, the multi-matching problem is difficult. It can be phrased as simultaneously solving multiple (NP-hard) quadratic assignment problems (QAPs) that are coupled via cycle-consistency constraints. The main limitations of existing multi-matching methods are that they either ignore geometric consistency and thus have limited robustness, or they are restricted to small-scale problems due to their (relatively) high computational cost. We address these shortcomings by introducing a Higher-order Projected Power Iteration method, which is (i) efficient and scales to tens of thousands of points, (ii) straightforward to implement, (iii) able to incorporate geometric consistency, and (iv) guarantees cycle-consistent multi-matchings. Experimentally we show that our approach is superior to existing methods.}, }
Endnote
%0 Report %A Bernard, Florian %A Thunberg, Johan %A Swoboda, Paul %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T Higher-order Projected Power Iterations for Scalable Multi-Matching : %G eng %U http://hdl.handle.net/21.11116/0000-0002-A8D0-5 %U http://arxiv.org/abs/1811.10541 %D 2018 %X The matching of multiple objects (e.g. shapes or images) is a fundamental problem in vision and graphics. In order to robustly handle ambiguities, noise and repetitive patterns in challenging real-world settings, it is essential to take geometric consistency between points into account. Computationally, the multi-matching problem is difficult. It can be phrased as simultaneously solving multiple (NP-hard) quadratic assignment problems (QAPs) that are coupled via cycle-consistency constraints. The main limitations of existing multi-matching methods are that they either ignore geometric consistency and thus have limited robustness, or they are restricted to small-scale problems due to their (relatively) high computational cost. We address these shortcomings by introducing a Higher-order Projected Power Iteration method, which is (i) efficient and scales to tens of thousands of points, (ii) straightforward to implement, (iii) able to incorporate geometric consistency, and (iv) guarantees cycle-consistent multi-matchings. Experimentally we show that our approach is superior to existing methods. %K Computer Science, Computer Vision and Pattern Recognition, cs.CV,Statistics, Machine Learning, stat.ML
Castelli Aleardi, L., Salihoglu, S., Singh, G., and Ovsjanikov, M. 2018. Spectral Measures of Distortion for Change Detection in Dynamic Graphs. https://hal.archives-ouvertes.fr/hal-01864079v2.
Export
BibTeX
@misc{Castelli_hal-01864079v2, TITLE = {Spectral Measures of Distortion for Change Detection in Dynamic Graphs}, AUTHOR = {Castelli Aleardi, Luca and Salihoglu, Semih and Singh, Gurprit and Ovsjanikov, Maks}, LANGUAGE = {eng}, URL = {https://hal.archives-ouvertes.fr/hal-01864079v2}, PUBLISHER = {HAL}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, }
Endnote
%0 Report %A Castelli Aleardi, Luca %A Salihoglu, Semih %A Singh, Gurprit %A Ovsjanikov, Maks %+ External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations %T Spectral Measures of Distortion for Change Detection in Dynamic Graphs : %G eng %U http://hdl.handle.net/21.11116/0000-0002-9F5D-4 %F OTHER: hal-01864079v2 %U https://hal.archives-ouvertes.fr/hal-01864079v2 %I HAL %D 2018 %U https://hal.archives-ouvertes.fr/hal-01864079/
Kim, H., Garrido, P., Tewari, A., et al. 2018c. Deep Video Portraits. http://arxiv.org/abs/1805.11714.
(arXiv: 1805.11714)
Abstract
We present a novel approach that enables photo-realistic re-animation of portrait videos using only an input video. In contrast to existing approaches that are restricted to manipulations of facial expressions only, we are the first to transfer the full 3D head position, head rotation, face expression, eye gaze, and eye blinking from a source actor to a portrait video of a target actor. The core of our approach is a generative neural network with a novel space-time architecture. The network takes as input synthetic renderings of a parametric face model, based on which it predicts photo-realistic video frames for a given target actor. The realism in this rendering-to-video transfer is achieved by careful adversarial training, and as a result, we can create modified target videos that mimic the behavior of the synthetically-created input. In order to enable source-to-target video re-animation, we render a synthetic target video with the reconstructed head animation parameters from a source video, and feed it into the trained network -- thus taking full control of the target. With the ability to freely recombine source and target parameters, we are able to demonstrate a large variety of video rewrite applications without explicitly modeling hair, body or background. For instance, we can reenact the full head using interactive user-controlled editing, and realize high-fidelity visual dubbing. To demonstrate the high quality of our output, we conduct an extensive series of experiments and evaluations, where for instance a user study shows that our video edits are hard to detect.
Export
BibTeX
@online{Kim_arXiv1805.11714, TITLE = {Deep Video Portraits}, AUTHOR = {Kim, Hyeongwoo and Garrido, Pablo and Tewari, Ayush and Xu, Weipeng and Thies, Justus and Nie{\ss}ner, Matthias and P{\'e}rez, Patrick and Richardt, Christian and Zollh{\"o}fer, Michael and Theobalt, Christian}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1805.11714}, EPRINT = {1805.11714}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {We present a novel approach that enables photo-realistic re-animation of portrait videos using only an input video. In contrast to existing approaches that are restricted to manipulations of facial expressions only, we are the first to transfer the full 3D head position, head rotation, face expression, eye gaze, and eye blinking from a source actor to a portrait video of a target actor. The core of our approach is a generative neural network with a novel space-time architecture. The network takes as input synthetic renderings of a parametric face model, based on which it predicts photo-realistic video frames for a given target actor. The realism in this rendering-to-video transfer is achieved by careful adversarial training, and as a result, we can create modified target videos that mimic the behavior of the synthetically-created input. In order to enable source-to-target video re-animation, we render a synthetic target video with the reconstructed head animation parameters from a source video, and feed it into the trained network -- thus taking full control of the target. With the ability to freely recombine source and target parameters, we are able to demonstrate a large variety of video rewrite applications without explicitly modeling hair, body or background. For instance, we can reenact the full head using interactive user-controlled editing, and realize high-fidelity visual dubbing. To demonstrate the high quality of our output, we conduct an extensive series of experiments and evaluations, where for instance a user study shows that our video edits are hard to detect.}, }
Endnote
%0 Report %A Kim, Hyeongwoo %A Garrido, Pablo %A Tewari, Ayush %A Xu, Weipeng %A Thies, Justus %A Nießner, Matthias %A Pérez, Patrick %A Richardt, Christian %A Zollhöfer, Michael %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations External Organizations External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T Deep Video Portraits : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5E1E-5 %U http://arxiv.org/abs/1805.11714 %D 2018 %X We present a novel approach that enables photo-realistic re-animation of portrait videos using only an input video. In contrast to existing approaches that are restricted to manipulations of facial expressions only, we are the first to transfer the full 3D head position, head rotation, face expression, eye gaze, and eye blinking from a source actor to a portrait video of a target actor. The core of our approach is a generative neural network with a novel space-time architecture. The network takes as input synthetic renderings of a parametric face model, based on which it predicts photo-realistic video frames for a given target actor. The realism in this rendering-to-video transfer is achieved by careful adversarial training, and as a result, we can create modified target videos that mimic the behavior of the synthetically-created input. In order to enable source-to-target video re-animation, we render a synthetic target video with the reconstructed head animation parameters from a source video, and feed it into the trained network -- thus taking full control of the target. With the ability to freely recombine source and target parameters, we are able to demonstrate a large variety of video rewrite applications without explicitly modeling hair, body or background. For instance, we can reenact the full head using interactive user-controlled editing, and realize high-fidelity visual dubbing. To demonstrate the high quality of our output, we conduct an extensive series of experiments and evaluations, where for instance a user study shows that our video edits are hard to detect. %K Computer Science, Computer Vision and Pattern Recognition, cs.CV,Computer Science, Artificial Intelligence, cs.AI,Computer Science, Graphics, cs.GR
Leimkühler, T., Singh, G., Myszkowski, K., Seidel, H.-P., and Ritschel, T. 2018c. End-to-end Sampling Patterns. http://arxiv.org/abs/1806.06710.
(arXiv: 1806.06710)
Abstract
Sample patterns have many uses in Computer Graphics, ranging from procedural object placement over Monte Carlo image synthesis to non-photorealistic depiction. Their properties such as discrepancy, spectra, anisotropy, or progressiveness have been analyzed extensively. However, designing methods to produce sampling patterns with certain properties can require substantial hand-crafting effort, both in coding, mathematical derivation and compute time. In particular, there is no systematic way to derive the best sampling algorithm for a specific end-task. Tackling this issue, we suggest another level of abstraction: a toolkit to end-to-end optimize over all sampling methods to find the one producing user-prescribed properties such as discrepancy or a spectrum that best fit the end-task. A user simply implements the forward losses and the sampling method is found automatically -- without coding or mathematical derivation -- by making use of back-propagation abilities of modern deep learning frameworks. While this optimization takes long, at deployment time the sampling method is quick to execute as iterated unstructured non-linear filtering using radial basis functions (RBFs) to represent high-dimensional kernels. Several important previous methods are special cases of this approach, which we compare to previous work and demonstrate its usefulness in several typical Computer Graphics applications. Finally, we propose sampling patterns with properties not shown before, such as high-dimensional blue noise with projective properties.
Export
BibTeX
@online{Leimkuehler_arXiv1806.06710, TITLE = {End-to-end Sampling Patterns}, AUTHOR = {Leimk{\"u}hler, Thomas and Singh, Gurprit and Myszkowski, Karol and Seidel, Hans-Peter and Ritschel, Tobias}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1806.06710}, EPRINT = {1806.06710}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {Sample patterns have many uses in Computer Graphics, ranging from procedural object placement over Monte Carlo image synthesis to non-photorealistic depiction. Their properties such as discrepancy, spectra, anisotropy, or progressiveness have been analyzed extensively. However, designing methods to produce sampling patterns with certain properties can require substantial hand-crafting effort, both in coding, mathematical derivation and compute time. In particular, there is no systematic way to derive the best sampling algorithm for a specific end-task. Tackling this issue, we suggest another level of abstraction: a toolkit to end-to-end optimize over all sampling methods to find the one producing user-prescribed properties such as discrepancy or a spectrum that best fit the end-task. A user simply implements the forward losses and the sampling method is found automatically -- without coding or mathematical derivation -- by making use of back-propagation abilities of modern deep learning frameworks. While this optimization takes long, at deployment time the sampling method is quick to execute as iterated unstructured non-linear filtering using radial basis functions (RBFs) to represent high-dimensional kernels. Several important previous methods are special cases of this approach, which we compare to previous work and demonstrate its usefulness in several typical Computer Graphics applications. Finally, we propose sampling patterns with properties not shown before, such as high-dimensional blue noise with projective properties.}, }
Endnote
%0 Report %A Leimkühler, Thomas %A Singh, Gurprit %A Myszkowski, Karol %A Seidel, Hans-Peter %A Ritschel, Tobias %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations %T End-to-end Sampling Patterns : %G eng %U http://hdl.handle.net/21.11116/0000-0002-1376-4 %U http://arxiv.org/abs/1806.06710 %D 2018 %X Sample patterns have many uses in Computer Graphics, ranging from procedural object placement over Monte Carlo image synthesis to non-photorealistic depiction. Their properties such as discrepancy, spectra, anisotropy, or progressiveness have been analyzed extensively. However, designing methods to produce sampling patterns with certain properties can require substantial hand-crafting effort, both in coding, mathematical derivation and compute time. In particular, there is no systematic way to derive the best sampling algorithm for a specific end-task. Tackling this issue, we suggest another level of abstraction: a toolkit to end-to-end optimize over all sampling methods to find the one producing user-prescribed properties such as discrepancy or a spectrum that best fit the end-task. A user simply implements the forward losses and the sampling method is found automatically -- without coding or mathematical derivation -- by making use of back-propagation abilities of modern deep learning frameworks. While this optimization takes long, at deployment time the sampling method is quick to execute as iterated unstructured non-linear filtering using radial basis functions (RBFs) to represent high-dimensional kernels. Several important previous methods are special cases of this approach, which we compare to previous work and demonstrate its usefulness in several typical Computer Graphics applications. Finally, we propose sampling patterns with properties not shown before, such as high-dimensional blue noise with projective properties. %K Computer Science, Graphics, cs.GR
Lin, K.Z., Xu, W., Sun, Q., Theobalt, C., and Chua, T.-S. 2018. Learning a Disentangled Embedding for Monocular 3D Shape Retrieval and Pose Estimation. http://arxiv.org/abs/1812.09899.
(arXiv: 1812.09899)
Abstract
We propose a novel approach to jointly perform 3D object retrieval and pose estimation from monocular images.In order to make the method robust to real world scene variations in the images, e.g. texture, lighting and background,we learn an embedding space from 3D data that only includes the relevant information, namely the shape and pose.Our method can then be trained for robustness under real world scene variations without having to render a large training set simulating these variations. Our learned embedding explicitly disentangles a shape vector and a pose vector, which alleviates both pose bias for 3D shape retrieval and categorical bias for pose estimation. Having the learned disentangled embedding, we train a CNN to map the images to the embedding space, and then retrieve the closest 3D shape from the database and estimate the 6D pose of the object using the embedding vectors. Our method achieves 10.8 median error for pose estimation and 0.514 top-1-accuracy for category agnostic 3D object retrieval on the Pascal3D+ dataset. It therefore outperforms the previous state-of-the-art methods on both tasks.
Export
BibTeX
@online{Kyaw_arXiv1812.09899, TITLE = {{Learning a Disentangled Embedding for Monocular 3D Shape Retrieval and Pose Estimation}}, AUTHOR = {Lin, Kyaw Zaw and Xu, Weipeng and Sun, Qianru and Theobalt, Christian and Chua, Tat-Seng}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1812.09899}, EPRINT = {1812.09899}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {We propose a novel approach to jointly perform 3D object retrieval and pose estimation from monocular images.In order to make the method robust to real world scene variations in the images, e.g. texture, lighting and background,we learn an embedding space from 3D data that only includes the relevant information, namely the shape and pose.Our method can then be trained for robustness under real world scene variations without having to render a large training set simulating these variations. Our learned embedding explicitly disentangles a shape vector and a pose vector, which alleviates both pose bias for 3D shape retrieval and categorical bias for pose estimation. Having the learned disentangled embedding, we train a CNN to map the images to the embedding space, and then retrieve the closest 3D shape from the database and estimate the 6D pose of the object using the embedding vectors. Our method achieves 10.8 median error for pose estimation and 0.514 top-1-accuracy for category agnostic 3D object retrieval on the Pascal3D+ dataset. It therefore outperforms the previous state-of-the-art methods on both tasks.}, }
Endnote
%0 Report %A Lin, Kyaw Zaw %A Xu, Weipeng %A Sun, Qianru %A Theobalt, Christian %A Chua, Tat-Seng %+ External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations %T Learning a Disentangled Embedding for Monocular 3D Shape Retrieval and Pose Estimation : %G eng %U http://hdl.handle.net/21.11116/0000-0002-D519-2 %U http://arxiv.org/abs/1812.09899 %D 2018 %X We propose a novel approach to jointly perform 3D object retrieval and pose estimation from monocular images.In order to make the method robust to real world scene variations in the images, e.g. texture, lighting and background,we learn an embedding space from 3D data that only includes the relevant information, namely the shape and pose.Our method can then be trained for robustness under real world scene variations without having to render a large training set simulating these variations. Our learned embedding explicitly disentangles a shape vector and a pose vector, which alleviates both pose bias for 3D shape retrieval and categorical bias for pose estimation. Having the learned disentangled embedding, we train a CNN to map the images to the embedding space, and then retrieve the closest 3D shape from the database and estimate the 6D pose of the object using the embedding vectors. Our method achieves 10.8 median error for pose estimation and 0.514 top-1-accuracy for category agnostic 3D object retrieval on the Pascal3D+ dataset. It therefore outperforms the previous state-of-the-art methods on both tasks. %K Computer Science, Computer Vision and Pattern Recognition, cs.CV
Liu, L., Xu, W., Zollhöfer, M., et al. 2018b. Neural Animation and Reenactment of Human Actor Videos. http://arxiv.org/abs/1809.03658.
(arXiv: 1809.03658)
Abstract
We propose a method for generating (near) video-realistic animations of real humans under user control. In contrast to conventional human character rendering, we do not require the availability of a production-quality photo-realistic 3D model of the human, but instead rely on a video sequence in conjunction with a (medium-quality) controllable 3D template model of the person. With that, our approach significantly reduces production cost compared to conventional rendering approaches based on production-quality 3D models, and can also be used to realistically edit existing videos. Technically, this is achieved by training a neural network that translates simple synthetic images of a human character into realistic imagery. For training our networks, we first track the 3D motion of the person in the video using the template model, and subsequently generate a synthetically rendered version of the video. These images are then used to train a conditional generative adversarial network that translates synthetic images of the 3D model into realistic imagery of the human. We evaluate our method for the reenactment of another person that is tracked in order to obtain the motion data, and show video results generated from artist-designed skeleton motion. Our results outperform the state-of-the-art in learning-based human image synthesis. Project page: http://gvv.mpi-inf.mpg.de/projects/wxu/HumanReenactment/
Export
BibTeX
@online{Liu_arXiv1809.03658, TITLE = {Neural Animation and Reenactment of Human Actor Videos}, AUTHOR = {Liu, Lingjie and Xu, Weipeng and Zollh{\"o}fer, Michael and Kim, Hyeongwoo and Bernard, Florian and Habermann, Marc and Wang, Wenping and Theobalt, Christian}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1809.03658}, EPRINT = {1809.03658}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {We propose a method for generating (near) video-realistic animations of real humans under user control. In contrast to conventional human character rendering, we do not require the availability of a production-quality photo-realistic 3D model of the human, but instead rely on a video sequence in conjunction with a (medium-quality) controllable 3D template model of the person. With that, our approach significantly reduces production cost compared to conventional rendering approaches based on production-quality 3D models, and can also be used to realistically edit existing videos. Technically, this is achieved by training a neural network that translates simple synthetic images of a human character into realistic imagery. For training our networks, we first track the 3D motion of the person in the video using the template model, and subsequently generate a synthetically rendered version of the video. These images are then used to train a conditional generative adversarial network that translates synthetic images of the 3D model into realistic imagery of the human. We evaluate our method for the reenactment of another person that is tracked in order to obtain the motion data, and show video results generated from artist-designed skeleton motion. Our results outperform the state-of-the-art in learning-based human image synthesis. Project page: http://gvv.mpi-inf.mpg.de/projects/wxu/HumanReenactment/}, }
Endnote
%0 Report %A Liu, Lingjie %A Xu, Weipeng %A Zollhöfer, Michael %A Kim, Hyeongwoo %A Bernard, Florian %A Habermann, Marc %A Wang, Wenping %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T Neural Animation and Reenactment of Human Actor Videos : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5E06-F %U http://arxiv.org/abs/1809.03658 %D 2018 %X We propose a method for generating (near) video-realistic animations of real humans under user control. In contrast to conventional human character rendering, we do not require the availability of a production-quality photo-realistic 3D model of the human, but instead rely on a video sequence in conjunction with a (medium-quality) controllable 3D template model of the person. With that, our approach significantly reduces production cost compared to conventional rendering approaches based on production-quality 3D models, and can also be used to realistically edit existing videos. Technically, this is achieved by training a neural network that translates simple synthetic images of a human character into realistic imagery. For training our networks, we first track the 3D motion of the person in the video using the template model, and subsequently generate a synthetically rendered version of the video. These images are then used to train a conditional generative adversarial network that translates synthetic images of the 3D model into realistic imagery of the human. We evaluate our method for the reenactment of another person that is tracked in order to obtain the motion data, and show video results generated from artist-designed skeleton motion. Our results outperform the state-of-the-art in learning-based human image synthesis. Project page: http://gvv.mpi-inf.mpg.de/projects/wxu/HumanReenactment/ %K Computer Science, Computer Vision and Pattern Recognition, cs.CV %U http://gvv.mpi-inf.mpg.de/projects/wxu/HumanReenactment/
Mehta, D., Kim, K.I., and Theobalt, C. 2018c. On Implicit Filter Level Sparsity in Convolutional Neural Networks. http://arxiv.org/abs/1811.12495.
(arXiv: 1811.12495)
Abstract
We investigate filter level sparsity that emerges in convolutional neural networks (CNNs) which employ Batch Normalization and ReLU activation, and are trained with adaptive gradient descent techniques and L2 regularization (or weight decay). We conduct an extensive experimental study casting these initial findings into hypotheses and conclusions about the mechanisms underlying the emergent filter level sparsity. This study allows new insight into the performance gap obeserved between adapative and non-adaptive gradient descent methods in practice. Further, analysis of the effect of training strategies and hyperparameters on the sparsity leads to practical suggestions in designing CNN training strategies enabling us to explore the tradeoffs between feature selectivity, network capacity, and generalization performance. Lastly, we show that the implicit sparsity can be harnessed for neural network speedup at par or better than explicit sparsification / pruning approaches, without needing any modifications to the typical training pipeline.
Export
BibTeX
@online{Mehta_arXIv1811.12495, TITLE = {On Implicit Filter Level Sparsity in Convolutional Neural Networks}, AUTHOR = {Mehta, Dushyant and Kim, Kwang In and Theobalt, Christian}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1811.12495}, EPRINT = {1811.12495}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {We investigate filter level sparsity that emerges in convolutional neural networks (CNNs) which employ Batch Normalization and ReLU activation, and are trained with adaptive gradient descent techniques and L2 regularization (or weight decay). We conduct an extensive experimental study casting these initial findings into hypotheses and conclusions about the mechanisms underlying the emergent filter level sparsity. This study allows new insight into the performance gap obeserved between adapative and non-adaptive gradient descent methods in practice. Further, analysis of the effect of training strategies and hyperparameters on the sparsity leads to practical suggestions in designing CNN training strategies enabling us to explore the tradeoffs between feature selectivity, network capacity, and generalization performance. Lastly, we show that the implicit sparsity can be harnessed for neural network speedup at par or better than explicit sparsification / pruning approaches, without needing any modifications to the typical training pipeline.}, }
Endnote
%0 Report %A Mehta, Dushyant %A Kim, Kwang In %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T On Implicit Filter Level Sparsity in Convolutional Neural Networks : %G eng %U http://hdl.handle.net/21.11116/0000-0002-E920-3 %U http://arxiv.org/abs/1811.12495 %D 2018 %X We investigate filter level sparsity that emerges in convolutional neural networks (CNNs) which employ Batch Normalization and ReLU activation, and are trained with adaptive gradient descent techniques and L2 regularization (or weight decay). We conduct an extensive experimental study casting these initial findings into hypotheses and conclusions about the mechanisms underlying the emergent filter level sparsity. This study allows new insight into the performance gap obeserved between adapative and non-adaptive gradient descent methods in practice. Further, analysis of the effect of training strategies and hyperparameters on the sparsity leads to practical suggestions in designing CNN training strategies enabling us to explore the tradeoffs between feature selectivity, network capacity, and generalization performance. Lastly, we show that the implicit sparsity can be harnessed for neural network speedup at par or better than explicit sparsification / pruning approaches, without needing any modifications to the typical training pipeline. %K Computer Science, Learning, cs.LG,Computer Science, Computer Vision and Pattern Recognition, cs.CV,eess.SP,Statistics, Machine Learning, stat.ML
Meka, A., Maximov, M., Zollhöfer, M., et al. 2018b. LIME: Live Intrinsic Material Estimation. http://arxiv.org/abs/1801.01075.
(arXiv: 1801.01075)
Abstract
We present the first end to end approach for real time material estimation for general object shapes with uniform material that only requires a single color image as input. In addition to Lambertian surface properties, our approach fully automatically computes the specular albedo, material shininess, and a foreground segmentation. We tackle this challenging and ill posed inverse rendering problem using recent advances in image to image translation techniques based on deep convolutional encoder decoder architectures. The underlying core representations of our approach are specular shading, diffuse shading and mirror images, which allow to learn the effective and accurate separation of diffuse and specular albedo. In addition, we propose a novel highly efficient perceptual rendering loss that mimics real world image formation and obtains intermediate results even during run time. The estimation of material parameters at real time frame rates enables exciting mixed reality applications, such as seamless illumination consistent integration of virtual objects into real world scenes, and virtual material cloning. We demonstrate our approach in a live setup, compare it to the state of the art, and demonstrate its effectiveness through quantitative and qualitative evaluation.
Export
BibTeX
@online{Meka_arXiv1801.01075, TITLE = {LIME: {L}ive Intrinsic Material Estimation}, AUTHOR = {Meka, Abhimitra and Maximov, Maxim and Zollh{\"o}fer, Michael and Chatterjee, Avishek and Seidel, Hans-Peter and Richardt, Christian and Theobalt, Christian}, URL = {http://arxiv.org/abs/1801.01075}, EPRINT = {1801.01075}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {We present the first end to end approach for real time material estimation for general object shapes with uniform material that only requires a single color image as input. In addition to Lambertian surface properties, our approach fully automatically computes the specular albedo, material shininess, and a foreground segmentation. We tackle this challenging and ill posed inverse rendering problem using recent advances in image to image translation techniques based on deep convolutional encoder decoder architectures. The underlying core representations of our approach are specular shading, diffuse shading and mirror images, which allow to learn the effective and accurate separation of diffuse and specular albedo. In addition, we propose a novel highly efficient perceptual rendering loss that mimics real world image formation and obtains intermediate results even during run time. The estimation of material parameters at real time frame rates enables exciting mixed reality applications, such as seamless illumination consistent integration of virtual objects into real world scenes, and virtual material cloning. We demonstrate our approach in a live setup, compare it to the state of the art, and demonstrate its effectiveness through quantitative and qualitative evaluation.}, }
Endnote
%0 Report %A Meka, Abhimitra %A Maximov, Maxim %A Zollhöfer, Michael %A Chatterjee, Avishek %A Seidel, Hans-Peter %A Richardt, Christian %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society D2 External Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T LIME: Live Intrinsic Material Estimation : %U http://hdl.handle.net/21.11116/0000-0001-40D9-2 %U http://arxiv.org/abs/1801.01075 %D 2018 %X We present the first end to end approach for real time material estimation for general object shapes with uniform material that only requires a single color image as input. In addition to Lambertian surface properties, our approach fully automatically computes the specular albedo, material shininess, and a foreground segmentation. We tackle this challenging and ill posed inverse rendering problem using recent advances in image to image translation techniques based on deep convolutional encoder decoder architectures. The underlying core representations of our approach are specular shading, diffuse shading and mirror images, which allow to learn the effective and accurate separation of diffuse and specular albedo. In addition, we propose a novel highly efficient perceptual rendering loss that mimics real world image formation and obtains intermediate results even during run time. The estimation of material parameters at real time frame rates enables exciting mixed reality applications, such as seamless illumination consistent integration of virtual objects into real world scenes, and virtual material cloning. We demonstrate our approach in a live setup, compare it to the state of the art, and demonstrate its effectiveness through quantitative and qualitative evaluation. %K Computer Science, Computer Vision and Pattern Recognition, cs.CV
Mlakar, D., Winter, M., Seidel, H.-P., Steinberger, M., and Zayer, R. 2018. AlSub: Fully Parallel and Modular Subdivision. http://arxiv.org/abs/1809.06047.
(arXiv: 1809.06047)
Abstract
In recent years, mesh subdivision---the process of forging smooth free-form surfaces from coarse polygonal meshes---has become an indispensable production instrument. Although subdivision performance is crucial during simulation, animation and rendering, state-of-the-art approaches still rely on serial implementations for complex parts of the subdivision process. Therefore, they often fail to harness the power of modern parallel devices, like the graphics processing unit (GPU), for large parts of the algorithm and must resort to time-consuming serial preprocessing. In this paper, we show that a complete parallelization of the subdivision process for modern architectures is possible. Building on sparse matrix linear algebra, we show how to structure the complete subdivision process into a sequence of algebra operations. By restructuring and grouping these operations, we adapt the process for different use cases, such as regular subdivision of dynamic meshes, uniform subdivision for immutable topology, and feature-adaptive subdivision for efficient rendering of animated models. As the same machinery is used for all use cases, identical subdivision results are achieved in all parts of the production pipeline. As a second contribution, we show how these linear algebra formulations can effectively be translated into efficient GPU kernels. Applying our strategies to $\sqrt{3}$, Loop and Catmull-Clark subdivision shows significant speedups of our approach compared to state-of-the-art solutions, while we completely avoid serial preprocessing.
Export
BibTeX
@online{Mlakar_arXiv1809.06047, TITLE = {{AlSub}: {Fully Parallel and Modular Subdivision}}, AUTHOR = {Mlakar, Daniel and Winter, Martin and Seidel, Hans-Peter and Steinberger, Markus and Zayer, Rhaleb}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1809.06047}, EPRINT = {1809.06047}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {In recent years, mesh subdivision---the process of forging smooth free-form surfaces from coarse polygonal meshes---has become an indispensable production instrument. Although subdivision performance is crucial during simulation, animation and rendering, state-of-the-art approaches still rely on serial implementations for complex parts of the subdivision process. Therefore, they often fail to harness the power of modern parallel devices, like the graphics processing unit (GPU), for large parts of the algorithm and must resort to time-consuming serial preprocessing. In this paper, we show that a complete parallelization of the subdivision process for modern architectures is possible. Building on sparse matrix linear algebra, we show how to structure the complete subdivision process into a sequence of algebra operations. By restructuring and grouping these operations, we adapt the process for different use cases, such as regular subdivision of dynamic meshes, uniform subdivision for immutable topology, and feature-adaptive subdivision for efficient rendering of animated models. As the same machinery is used for all use cases, identical subdivision results are achieved in all parts of the production pipeline. As a second contribution, we show how these linear algebra formulations can effectively be translated into efficient GPU kernels. Applying our strategies to $\sqrt{3}$, Loop and Catmull-Clark subdivision shows significant speedups of our approach compared to state-of-the-art solutions, while we completely avoid serial preprocessing.}, }
Endnote
%0 Report %A Mlakar, Daniel %A Winter, Martin %A Seidel, Hans-Peter %A Steinberger, Markus %A Zayer, Rhaleb %+ External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T AlSub: Fully Parallel and Modular Subdivision : %G eng %U http://hdl.handle.net/21.11116/0000-0002-E5E2-C %U http://arxiv.org/abs/1809.06047 %D 2018 %X In recent years, mesh subdivision---the process of forging smooth free-form surfaces from coarse polygonal meshes---has become an indispensable production instrument. Although subdivision performance is crucial during simulation, animation and rendering, state-of-the-art approaches still rely on serial implementations for complex parts of the subdivision process. Therefore, they often fail to harness the power of modern parallel devices, like the graphics processing unit (GPU), for large parts of the algorithm and must resort to time-consuming serial preprocessing. In this paper, we show that a complete parallelization of the subdivision process for modern architectures is possible. Building on sparse matrix linear algebra, we show how to structure the complete subdivision process into a sequence of algebra operations. By restructuring and grouping these operations, we adapt the process for different use cases, such as regular subdivision of dynamic meshes, uniform subdivision for immutable topology, and feature-adaptive subdivision for efficient rendering of animated models. As the same machinery is used for all use cases, identical subdivision results are achieved in all parts of the production pipeline. As a second contribution, we show how these linear algebra formulations can effectively be translated into efficient GPU kernels. Applying our strategies to $\sqrt{3}$, Loop and Catmull-Clark subdivision shows significant speedups of our approach compared to state-of-the-art solutions, while we completely avoid serial preprocessing. %K Computer Science, Graphics, cs.GR
Serrano, A., Gutierrez, D., Myszkowski, K., Seidel, H.-P., and Masia, B. 2018. An Intuitive Control Space for Material Appearance. http://arxiv.org/abs/1806.04950.
(arXiv: 1806.04950)
Abstract
Many different techniques for measuring material appearance have been proposed in the last few years. These have produced large public datasets, which have been used for accurate, data-driven appearance modeling. However, although these datasets have allowed us to reach an unprecedented level of realism in visual appearance, editing the captured data remains a challenge. In this paper, we present an intuitive control space for predictable editing of captured BRDF data, which allows for artistic creation of plausible novel material appearances, bypassing the difficulty of acquiring novel samples. We first synthesize novel materials, extending the existing MERL dataset up to 400 mathematically valid BRDFs. We then design a large-scale experiment, gathering 56,000 subjective ratings on the high-level perceptual attributes that best describe our extended dataset of materials. Using these ratings, we build and train networks of radial basis functions to act as functionals mapping the perceptual attributes to an underlying PCA-based representation of BRDFs. We show that our functionals are excellent predictors of the perceived attributes of appearance. Our control space enables many applications, including intuitive material editing of a wide range of visual properties, guidance for gamut mapping, analysis of the correlation between perceptual attributes, or novel appearance similarity metrics. Moreover, our methodology can be used to derive functionals applicable to classic analytic BRDF representations. We release our code and dataset publicly, in order to support and encourage further research in this direction.
Export
BibTeX
@online{Serrano_arXiv1806.04950, TITLE = {An Intuitive Control Space for Material Appearance}, AUTHOR = {Serrano, Ana and Gutierrez, Diego and Myszkowski, Karol and Seidel, Hans-Peter and Masia, Belen}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1806.04950}, EPRINT = {1806.04950}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {Many different techniques for measuring material appearance have been proposed in the last few years. These have produced large public datasets, which have been used for accurate, data-driven appearance modeling. However, although these datasets have allowed us to reach an unprecedented level of realism in visual appearance, editing the captured data remains a challenge. In this paper, we present an intuitive control space for predictable editing of captured BRDF data, which allows for artistic creation of plausible novel material appearances, bypassing the difficulty of acquiring novel samples. We first synthesize novel materials, extending the existing MERL dataset up to 400 mathematically valid BRDFs. We then design a large-scale experiment, gathering 56,000 subjective ratings on the high-level perceptual attributes that best describe our extended dataset of materials. Using these ratings, we build and train networks of radial basis functions to act as functionals mapping the perceptual attributes to an underlying PCA-based representation of BRDFs. We show that our functionals are excellent predictors of the perceived attributes of appearance. Our control space enables many applications, including intuitive material editing of a wide range of visual properties, guidance for gamut mapping, analysis of the correlation between perceptual attributes, or novel appearance similarity metrics. Moreover, our methodology can be used to derive functionals applicable to classic analytic BRDF representations. We release our code and dataset publicly, in order to support and encourage further research in this direction.}, }
Endnote
%0 Report %A Serrano, Ana %A Gutierrez, Diego %A Myszkowski, Karol %A Seidel, Hans-Peter %A Masia, Belen %+ External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations %T An Intuitive Control Space for Material Appearance : %G eng %U http://hdl.handle.net/21.11116/0000-0002-151E-6 %U http://arxiv.org/abs/1806.04950 %D 2018 %X Many different techniques for measuring material appearance have been proposed in the last few years. These have produced large public datasets, which have been used for accurate, data-driven appearance modeling. However, although these datasets have allowed us to reach an unprecedented level of realism in visual appearance, editing the captured data remains a challenge. In this paper, we present an intuitive control space for predictable editing of captured BRDF data, which allows for artistic creation of plausible novel material appearances, bypassing the difficulty of acquiring novel samples. We first synthesize novel materials, extending the existing MERL dataset up to 400 mathematically valid BRDFs. We then design a large-scale experiment, gathering 56,000 subjective ratings on the high-level perceptual attributes that best describe our extended dataset of materials. Using these ratings, we build and train networks of radial basis functions to act as functionals mapping the perceptual attributes to an underlying PCA-based representation of BRDFs. We show that our functionals are excellent predictors of the perceived attributes of appearance. Our control space enables many applications, including intuitive material editing of a wide range of visual properties, guidance for gamut mapping, analysis of the correlation between perceptual attributes, or novel appearance similarity metrics. Moreover, our methodology can be used to derive functionals applicable to classic analytic BRDF representations. We release our code and dataset publicly, in order to support and encourage further research in this direction. %K Computer Science, Graphics, cs.GR
Sun, Q., Tewari, A., Xu, W., Fritz, M., Theobalt, C., and Schiele, B. 2018b. A Hybrid Model for Identity Obfuscation by Face Replacement. http://arxiv.org/abs/1804.04779.
(arXiv: 1804.04779)
Abstract
As more and more personal photos are shared and tagged in social media, avoiding privacy risks such as unintended recognition becomes increasingly challenging. We propose a new hybrid approach to obfuscate identities in photos by head replacement. Our approach combines state of the art parametric face synthesis with latest advances in Generative Adversarial Networks (GAN) for data-driven image synthesis. On the one hand, the parametric part of our method gives us control over the facial parameters and allows for explicit manipulation of the identity. On the other hand, the data-driven aspects allow for adding fine details and overall realism as well as seamless blending into the scene context. In our experiments, we show highly realistic output of our system that improves over the previous state of the art in obfuscation rate while preserving a higher similarity to the original image content.
Export
BibTeX
@online{Sun_arXiv1804.04779, TITLE = {A Hybrid Model for Identity Obfuscation by Face Replacement}, AUTHOR = {Sun, Qianru and Tewari, Ayush and Xu, Weipeng and Fritz, Mario and Theobalt, Christian and Schiele, Bernt}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1804.04779}, EPRINT = {1804.04779}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {As more and more personal photos are shared and tagged in social media, avoiding privacy risks such as unintended recognition becomes increasingly challenging. We propose a new hybrid approach to obfuscate identities in photos by head replacement. Our approach combines state of the art parametric face synthesis with latest advances in Generative Adversarial Networks (GAN) for data-driven image synthesis. On the one hand, the parametric part of our method gives us control over the facial parameters and allows for explicit manipulation of the identity. On the other hand, the data-driven aspects allow for adding fine details and overall realism as well as seamless blending into the scene context. In our experiments, we show highly realistic output of our system that improves over the previous state of the art in obfuscation rate while preserving a higher similarity to the original image content.}, }
Endnote
%0 Report %A Sun, Qianru %A Tewari, Ayush %A Xu, Weipeng %A Fritz, Mario %A Theobalt, Christian %A Schiele, Bernt %+ Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Vision and Multimodal Computing, MPI for Informatics, Max Planck Society %T A Hybrid Model for Identity Obfuscation by Face Replacement : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5E25-C %U http://arxiv.org/abs/1804.04779 %D 2018 %X As more and more personal photos are shared and tagged in social media, avoiding privacy risks such as unintended recognition becomes increasingly challenging. We propose a new hybrid approach to obfuscate identities in photos by head replacement. Our approach combines state of the art parametric face synthesis with latest advances in Generative Adversarial Networks (GAN) for data-driven image synthesis. On the one hand, the parametric part of our method gives us control over the facial parameters and allows for explicit manipulation of the identity. On the other hand, the data-driven aspects allow for adding fine details and overall realism as well as seamless blending into the scene context. In our experiments, we show highly realistic output of our system that improves over the previous state of the art in obfuscation rate while preserving a higher similarity to the original image content. %K Computer Science, Computer Vision and Pattern Recognition, cs.CV,Computer Science, Cryptography and Security, cs.CR
Tewari, A., Bernard, F., Garrido, P., et al. 2018c. FML: Face Model Learning from Videos. http://arxiv.org/abs/1812.07603.
(arXiv: 1812.07603)
Abstract
Monocular image-based 3D reconstruction of faces is a long-standing problem in computer vision. Since image data is a 2D projection of a 3D face, the resulting depth ambiguity makes the problem ill-posed. Most existing methods rely on data-driven priors that are built from limited 3D face scans. In contrast, we propose multi-frame video-based self-supervised training of a deep network that (i) learns a face identity model both in shape and appearance while (ii) jointly learning to reconstruct 3D faces. Our face model is learned using only corpora of in-the-wild video clips collected from the Internet. This virtually endless source of training data enables learning of a highly general 3D face model. In order to achieve this, we propose a novel multi-frame consistency loss that ensures consistent shape and appearance across multiple frames of a subject's face, thus minimizing depth ambiguity. At test time we can use an arbitrary number of frames, so that we can perform both monocular as well as multi-frame reconstruction.
Export
BibTeX
@online{tewari2018fml, TITLE = {{FML}: {Face Model Learning from Videos}}, AUTHOR = {Tewari, Ayush and Bernard, Florian and Garrido, Pablo and Bharaj, Gaurav and Elgharib, Mohamed and Seidel, Hans-Peter and P{\'e}rez, Patrick and Zollh{\"o}fer, Michael and Theobalt, Christian}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1812.07603}, EPRINT = {1812.07603}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {Monocular image-based 3D reconstruction of faces is a long-standing problem in computer vision. Since image data is a 2D projection of a 3D face, the resulting depth ambiguity makes the problem ill-posed. Most existing methods rely on data-driven priors that are built from limited 3D face scans. In contrast, we propose multi-frame video-based self-supervised training of a deep network that (i) learns a face identity model both in shape and appearance while (ii) jointly learning to reconstruct 3D faces. Our face model is learned using only corpora of in-the-wild video clips collected from the Internet. This virtually endless source of training data enables learning of a highly general 3D face model. In order to achieve this, we propose a novel multi-frame consistency loss that ensures consistent shape and appearance across multiple frames of a subject's face, thus minimizing depth ambiguity. At test time we can use an arbitrary number of frames, so that we can perform both monocular as well as multi-frame reconstruction.}, }
Endnote
%0 Report %A Tewari, Ayush %A Bernard, Florian %A Garrido, Pablo %A Bharaj, Gaurav %A Elgharib, Mohamed %A Seidel, Hans-Peter %A Pérez, Patrick %A Zollhöfer, Michael %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T FML: Face Model Learning from Videos : %G eng %U http://hdl.handle.net/21.11116/0000-0002-EF79-A %U http://arxiv.org/abs/1812.07603 %D 2018 %X Monocular image-based 3D reconstruction of faces is a long-standing problem in computer vision. Since image data is a 2D projection of a 3D face, the resulting depth ambiguity makes the problem ill-posed. Most existing methods rely on data-driven priors that are built from limited 3D face scans. In contrast, we propose multi-frame video-based self-supervised training of a deep network that (i) learns a face identity model both in shape and appearance while (ii) jointly learning to reconstruct 3D faces. Our face model is learned using only corpora of in-the-wild video clips collected from the Internet. This virtually endless source of training data enables learning of a highly general 3D face model. In order to achieve this, we propose a novel multi-frame consistency loss that ensures consistent shape and appearance across multiple frames of a subject's face, thus minimizing depth ambiguity. At test time we can use an arbitrary number of frames, so that we can perform both monocular as well as multi-frame reconstruction. %K Computer Science, Computer Vision and Pattern Recognition, cs.CV %U https://www.youtube.com/watch?v=SG2BwxCw0lQ
Thies, J., Zollhöfer, M., Theobalt, C., Stamminger, M., and Nießner, M. 2018c. HeadOn: Real-time Reenactment of Human Portrait Videos. http://arxiv.org/abs/1805.11729.
(arXiv: 1805.11729)
Abstract
We propose HeadOn, the first real-time source-to-target reenactment approach for complete human portrait videos that enables transfer of torso and head motion, face expression, and eye gaze. Given a short RGB-D video of the target actor, we automatically construct a personalized geometry proxy that embeds a parametric head, eye, and kinematic torso model. A novel real-time reenactment algorithm employs this proxy to photo-realistically map the captured motion from the source actor to the target actor. On top of the coarse geometric proxy, we propose a video-based rendering technique that composites the modified target portrait video via view- and pose-dependent texturing, and creates photo-realistic imagery of the target actor under novel torso and head poses, facial expressions, and gaze directions. To this end, we propose a robust tracking of the face and torso of the source actor. We extensively evaluate our approach and show significant improvements in enabling much greater flexibility in creating realistic reenacted output videos.
Export
BibTeX
@online{Thies_arXiv1805.11729, TITLE = {{HeadOn}: {R}eal-time Reenactment of Human Portrait Videos}, AUTHOR = {Thies, Justus and Zollh{\"o}fer, Michael and Theobalt, Christian and Stamminger, Marc and Nie{\ss}ner, Matthias}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1805.11729}, EPRINT = {1805.11729}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {We propose HeadOn, the first real-time source-to-target reenactment approach for complete human portrait videos that enables transfer of torso and head motion, face expression, and eye gaze. Given a short RGB-D video of the target actor, we automatically construct a personalized geometry proxy that embeds a parametric head, eye, and kinematic torso model. A novel real-time reenactment algorithm employs this proxy to photo-realistically map the captured motion from the source actor to the target actor. On top of the coarse geometric proxy, we propose a video-based rendering technique that composites the modified target portrait video via view- and pose-dependent texturing, and creates photo-realistic imagery of the target actor under novel torso and head poses, facial expressions, and gaze directions. To this end, we propose a robust tracking of the face and torso of the source actor. We extensively evaluate our approach and show significant improvements in enabling much greater flexibility in creating realistic reenacted output videos.}, }
Endnote
%0 Report %A Thies, Justus %A Zollhöfer, Michael %A Theobalt, Christian %A Stamminger, Marc %A Nießner, Matthias %+ External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations %T HeadOn: Real-time Reenactment of Human Portrait Videos : %G eng %U http://hdl.handle.net/21.11116/0000-0002-5E14-F %U http://arxiv.org/abs/1805.11729 %D 2018 %X We propose HeadOn, the first real-time source-to-target reenactment approach for complete human portrait videos that enables transfer of torso and head motion, face expression, and eye gaze. Given a short RGB-D video of the target actor, we automatically construct a personalized geometry proxy that embeds a parametric head, eye, and kinematic torso model. A novel real-time reenactment algorithm employs this proxy to photo-realistically map the captured motion from the source actor to the target actor. On top of the coarse geometric proxy, we propose a video-based rendering technique that composites the modified target portrait video via view- and pose-dependent texturing, and creates photo-realistic imagery of the target actor under novel torso and head poses, facial expressions, and gaze directions. To this end, we propose a robust tracking of the face and torso of the source actor. We extensively evaluate our approach and show significant improvements in enabling much greater flexibility in creating realistic reenacted output videos. %K Computer Science, Computer Vision and Pattern Recognition, cs.CV,Computer Science, Graphics, cs.GR
Thies, J., Zollhöfer, M., Theobalt, C., Stamminger, M., and Nießner, M. 2018d. IGNOR: Image-guided Neural Object Rendering. http://arxiv.org/abs/1811.10720.
(arXiv: 1811.10720)
Abstract
We propose a new learning-based novel view synthesis approach for scanned objects that is trained based on a set of multi-view images. Instead of using texture mapping or hand-designed image-based rendering, we directly train a deep neural network to synthesize a view-dependent image of an object. First, we employ a coverage-based nearest neighbour look-up to retrieve a set of reference frames that are explicitly warped to a given target view using cross-projection. Our network then learns to best composite the warped images. This enables us to generate photo-realistic results, while not having to allocate capacity on `remembering' object appearance. Instead, the multi-view images can be reused. While this works well for diffuse objects, cross-projection does not generalize to view-dependent effects. Therefore, we propose a decomposition network that extracts view-dependent effects and that is trained in a self-supervised manner. After decomposition, the diffuse shading is cross-projected, while the view-dependent layer of the target view is regressed. We show the effectiveness of our approach both qualitatively and quantitatively on real as well as synthetic data.
Export
BibTeX
@online{Thies2018IGNORIN, TITLE = {{IGNOR}: {Image-guided Neural Object Rendering}}, AUTHOR = {Thies, Justus and Zollh{\"o}fer, Michael and Theobalt, Christian and Stamminger, Marc and Nie{\ss}ner, Matthias}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1811.10720}, EPRINT = {1811.10720}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {We propose a new learning-based novel view synthesis approach for scanned objects that is trained based on a set of multi-view images. Instead of using texture mapping or hand-designed image-based rendering, we directly train a deep neural network to synthesize a view-dependent image of an object. First, we employ a coverage-based nearest neighbour look-up to retrieve a set of reference frames that are explicitly warped to a given target view using cross-projection. Our network then learns to best composite the warped images. This enables us to generate photo-realistic results, while not having to allocate capacity on `remembering' object appearance. Instead, the multi-view images can be reused. While this works well for diffuse objects, cross-projection does not generalize to view-dependent effects. Therefore, we propose a decomposition network that extracts view-dependent effects and that is trained in a self-supervised manner. After decomposition, the diffuse shading is cross-projected, while the view-dependent layer of the target view is regressed. We show the effectiveness of our approach both qualitatively and quantitatively on real as well as synthetic data.}, }
Endnote
%0 Report %A Thies, Justus %A Zollhöfer, Michael %A Theobalt, Christian %A Stamminger, Marc %A Nießner, Matthias %+ External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations %T IGNOR: Image-guided Neural Object Rendering : %G eng %U http://hdl.handle.net/21.11116/0000-0002-F7EB-F %U http://arxiv.org/abs/1811.10720 %D 2018 %X We propose a new learning-based novel view synthesis approach for scanned objects that is trained based on a set of multi-view images. Instead of using texture mapping or hand-designed image-based rendering, we directly train a deep neural network to synthesize a view-dependent image of an object. First, we employ a coverage-based nearest neighbour look-up to retrieve a set of reference frames that are explicitly warped to a given target view using cross-projection. Our network then learns to best composite the warped images. This enables us to generate photo-realistic results, while not having to allocate capacity on `remembering' object appearance. Instead, the multi-view images can be reused. While this works well for diffuse objects, cross-projection does not generalize to view-dependent effects. Therefore, we propose a decomposition network that extracts view-dependent effects and that is trained in a self-supervised manner. After decomposition, the diffuse shading is cross-projected, while the view-dependent layer of the target view is regressed. We show the effectiveness of our approach both qualitatively and quantitatively on real as well as synthetic data. %K Computer Science, Computer Vision and Pattern Recognition, cs.CV %U https://youtu.be/s79HG9yn7QM
Xu, W., Chatterjee, A., Zollhöfer, M., et al. 2018b. Mo2Cap2: Real-time Mobile 3D Motion Capture with a Cap-mounted Fisheye Camera. http://arxiv.org/abs/1803.05959.
(arXiv: 1803.05959)
Abstract
We propose the first real-time approach for the egocentric estimation of 3D human body pose in a wide range of unconstrained everyday activities. This setting has a unique set of challenges, such as mobility of the hardware setup, and robustness to long capture sessions with fast recovery from tracking failures. We tackle these challenges based on a novel lightweight setup that converts a standard baseball cap to a device for high-quality pose estimation based on a single cap-mounted fisheye camera. From the captured egocentric live stream, our CNN based 3D pose estimation approach runs at 60Hz on a consumer-level GPU. In addition to the novel hardware setup, our other main contributions are: 1) a large ground truth training corpus of top-down fisheye images and 2) a novel disentangled 3D pose estimation approach that takes the unique properties of the egocentric viewpoint into account. As shown by our evaluation, we achieve lower 3D joint error as well as better 2D overlay than the existing baselines.
Export
BibTeX
@online{Xu_arXiv1803.05959, TITLE = {{Mo2Cap2}: Real-time Mobile {3D} Motion Capture with a Cap-mounted Fisheye Camera}, AUTHOR = {Xu, Weipeng and Chatterjee, Avishek and Zollh{\"o}fer, Michael and Rhodin, Helge and Fua, Pascal and Seidel, Hans-Peter and Theobalt, Christian}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1803.05959}, EPRINT = {1803.05959}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {We propose the first real-time approach for the egocentric estimation of 3D human body pose in a wide range of unconstrained everyday activities. This setting has a unique set of challenges, such as mobility of the hardware setup, and robustness to long capture sessions with fast recovery from tracking failures. We tackle these challenges based on a novel lightweight setup that converts a standard baseball cap to a device for high-quality pose estimation based on a single cap-mounted fisheye camera. From the captured egocentric live stream, our CNN based 3D pose estimation approach runs at 60Hz on a consumer-level GPU. In addition to the novel hardware setup, our other main contributions are: 1) a large ground truth training corpus of top-down fisheye images and 2) a novel disentangled 3D pose estimation approach that takes the unique properties of the egocentric viewpoint into account. As shown by our evaluation, we achieve lower 3D joint error as well as better 2D overlay than the existing baselines.}, }
Endnote
%0 Report %A Xu, Weipeng %A Chatterjee, Avishek %A Zollhöfer, Michael %A Rhodin, Helge %A Fua, Pascal %A Seidel, Hans-Peter %A Theobalt, Christian %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T Mo2Cap2: Real-time Mobile 3D Motion Capture with a Cap-mounted Fisheye Camera : %G eng %U http://hdl.handle.net/21.11116/0000-0001-3C65-B %U http://arxiv.org/abs/1803.05959 %D 2018 %X We propose the first real-time approach for the egocentric estimation of 3D human body pose in a wide range of unconstrained everyday activities. This setting has a unique set of challenges, such as mobility of the hardware setup, and robustness to long capture sessions with fast recovery from tracking failures. We tackle these challenges based on a novel lightweight setup that converts a standard baseball cap to a device for high-quality pose estimation based on a single cap-mounted fisheye camera. From the captured egocentric live stream, our CNN based 3D pose estimation approach runs at 60Hz on a consumer-level GPU. In addition to the novel hardware setup, our other main contributions are: 1) a large ground truth training corpus of top-down fisheye images and 2) a novel disentangled 3D pose estimation approach that takes the unique properties of the egocentric viewpoint into account. As shown by our evaluation, we achieve lower 3D joint error as well as better 2D overlay than the existing baselines. %K Computer Science, Computer Vision and Pattern Recognition, cs.CV
Zayer, R., Mlakar, D., Steinberger, M., and Seidel, H.-P. 2018b. Layered Fields for Natural Tessellations on Surfaces. http://arxiv.org/abs/1804.09152.
(arXiv: 1804.09152)
Abstract
Mimicking natural tessellation patterns is a fascinating multi-disciplinary problem. Geometric methods aiming at reproducing such partitions on surface meshes are commonly based on the Voronoi model and its variants, and are often faced with challenging issues such as metric estimation, geometric, topological complications, and most critically parallelization. In this paper, we introduce an alternate model which may be of value for resolving these issues. We drop the assumption that regions need to be separated by lines. Instead, we regard region boundaries as narrow bands and we model the partition as a set of smooth functions layered over the surface. Given an initial set of seeds or regions, the partition emerges as the solution of a time dependent set of partial differential equations describing concurrently evolving fronts on the surface. Our solution does not require geodesic estimation, elaborate numerical solvers, or complicated bookkeeping data structures. The cost per time-iteration is dominated by the multiplication and addition of two sparse matrices. Extension of our approach in a Lloyd's algorithm fashion can be easily achieved and the extraction of the dual mesh can be conveniently preformed in parallel through matrix algebra. As our approach relies mainly on basic linear algebra kernels, it lends itself to efficient implementation on modern graphics hardware.
Export
BibTeX
@online{Zayer_arXiv1804.09152, TITLE = {Layered Fields for Natural Tessellations on Surfaces}, AUTHOR = {Zayer, Rhaleb and Mlakar, Daniel and Steinberger, Markus and Seidel, Hans-Peter}, LANGUAGE = {eng}, URL = {http://arxiv.org/abs/1804.09152}, EPRINT = {1804.09152}, EPRINTTYPE = {arXiv}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, ABSTRACT = {Mimicking natural tessellation patterns is a fascinating multi-disciplinary problem. Geometric methods aiming at reproducing such partitions on surface meshes are commonly based on the Voronoi model and its variants, and are often faced with challenging issues such as metric estimation, geometric, topological complications, and most critically parallelization. In this paper, we introduce an alternate model which may be of value for resolving these issues. We drop the assumption that regions need to be separated by lines. Instead, we regard region boundaries as narrow bands and we model the partition as a set of smooth functions layered over the surface. Given an initial set of seeds or regions, the partition emerges as the solution of a time dependent set of partial differential equations describing concurrently evolving fronts on the surface. Our solution does not require geodesic estimation, elaborate numerical solvers, or complicated bookkeeping data structures. The cost per time-iteration is dominated by the multiplication and addition of two sparse matrices. Extension of our approach in a Lloyd's algorithm fashion can be easily achieved and the extraction of the dual mesh can be conveniently preformed in parallel through matrix algebra. As our approach relies mainly on basic linear algebra kernels, it lends itself to efficient implementation on modern graphics hardware.}, }
Endnote
%0 Report %A Zayer, Rhaleb %A Mlakar, Daniel %A Steinberger, Markus %A Seidel, Hans-Peter %+ Computer Graphics, MPI for Informatics, Max Planck Society External Organizations External Organizations Computer Graphics, MPI for Informatics, Max Planck Society %T Layered Fields for Natural Tessellations on Surfaces : %G eng %U http://hdl.handle.net/21.11116/0000-0002-152D-5 %U http://arxiv.org/abs/1804.09152 %D 2018 %X Mimicking natural tessellation patterns is a fascinating multi-disciplinary problem. Geometric methods aiming at reproducing such partitions on surface meshes are commonly based on the Voronoi model and its variants, and are often faced with challenging issues such as metric estimation, geometric, topological complications, and most critically parallelization. In this paper, we introduce an alternate model which may be of value for resolving these issues. We drop the assumption that regions need to be separated by lines. Instead, we regard region boundaries as narrow bands and we model the partition as a set of smooth functions layered over the surface. Given an initial set of seeds or regions, the partition emerges as the solution of a time dependent set of partial differential equations describing concurrently evolving fronts on the surface. Our solution does not require geodesic estimation, elaborate numerical solvers, or complicated bookkeeping data structures. The cost per time-iteration is dominated by the multiplication and addition of two sparse matrices. Extension of our approach in a Lloyd's algorithm fashion can be easily achieved and the extraction of the dual mesh can be conveniently preformed in parallel through matrix algebra. As our approach relies mainly on basic linear algebra kernels, it lends itself to efficient implementation on modern graphics hardware. %K Computer Science, Graphics, cs.GR,Computer Science, Distributed, Parallel, and Cluster Computing, cs.DC
Thesis
Hajipour, H. 2018. Weakly-supervised Surface Reconstruction Using Floating Radial Basis Functions. .
Export
BibTeX
@mastersthesis{HajipourMSc2018, TITLE = {Weakly-supervised Surface Reconstruction Using Floating Radial Basis Functions}, AUTHOR = {Hajipour, Hossein}, LANGUAGE = {eng}, SCHOOL = {Universit{\"a}t des Saarlandes}, ADDRESS = {Saarbr{\"u}cken}, YEAR = {2018}, MARGINALMARK = {$\bullet$}, DATE = {2018}, }
Endnote
%0 Thesis %A Hajipour, Hossein %Y Theobalt, Christian %A referee: Tewari, Ayush %+ Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society Computer Graphics, MPI for Informatics, Max Planck Society %T Weakly-supervised Surface Reconstruction Using Floating Radial Basis Functions : %G eng %U http://hdl.handle.net/21.11116/0000-0003-E013-A %I Universität des Saarlandes %C Saarbrücken %D 2018 %V master %9 master