MPII Cooking Activities Dataset

This site hosts the MPII (Max Planck Institute for Informatics) Cooking Activities dataset as well the corresponding publications [1] and code.

Please contact us if you have questions

Change log

• 21/01/14: added MATLAB code for sift-based part tracker
• 13/09/26: updated pose challenge (version 1.1): Test frames now contain ids where to find them in full dataset.
• 13/09/25: added continous pose challenge
• 13/09/20: added 3D pose challenge [2]
• 13/09: added detection results for download

Download

License

The data is only to be used for scientific purposes and must not be republished other than by the Max Planck Institute for Informatics. The scientific use includes processing the data and showing it in publications and presentations. When using it please cite [1].

Activity Recognition dataset

• README
• Raw data
  • videos as avis (8.7 GB)
  • images as jpgs (133 GB)
• Video features
  • dense trajectory features (raw features 391GB)
    • trajectory (4.0 GB)
    • hog (4.1 GB)
    • hof (3.9 GB)
    • mbh (4.1 GB)
  • pose features
    • bm (1.6 GB)
    • fft (1.2 GB)
  • intermediate pose features (pose, tracks, 138 MB)

Evaluation Code and data

• code (version 1.0)
• data (version 1.0): includes our classification and detection results to produce the tables in the paper
• ground truth annotations

Pose challenge

• Images, annotations, and evaluation code (version 1.2, 360 MB)
• 3D pose challenge[2]

Continous pose challenge

Continous frames labeled with human pose (no published results on this so far).

• Images, annotations, and evaluation code (751 MB)

Matlab Code for SIFT based Part Tracker

• code

We highly recommend a download manager which allows to continue interrupted downloads, under Linux you could try wget. 

Acknowledgements

The authors would like to thank the annotators for the effort they put into this project, among others:
Seyed Mehdi Khodadad Hosseini, and Dragana Majstorovic

Hardware

The dataset was recorded with a camera system from 4D View Solutions.

Help & Contact

If you need any help or have any suggestions feel free to contact Marcus Rohrbach.
We are looking forward to hear from your success of using the dataset.

Feel free to subsribe to our mailing list to get updates (cookingactivities-join@lists.mpi-inf.mpg.de).

Related datasets of our group

• MPII Cooking Composite Activities (a superset focusing on classification of composed activites)
• Saarbrücken Corpus of Textually Annotated Cooking Scenes (short: TACoS) textual descriptions for MPII Cooking Composite Activities
• MPII Multi-Kinect Dataset: Multi-view kinect object classification, recorded in the same kitchen