Weakly Supervised Object Boundaries

Anna Khoreva, Rodrigo Benenson, Mohamed Omran,

Matthias Hein and Bernt Schiele

Abstract

State-of-the-art learning based boundary detection methods require extensive training data. Since labelling object boundaries is one of the most expensive types of annotations, there is a need to relax the requirement to carefully annotate images to make both the training more affordable and to extend the amount of training data. In this paper we propose a technique to generate weakly supervised annotations and show that bounding box annotations alone suffice to reach high-quality object boundaries without using any object-specific boundary annotations. With the proposed weak supervision techniques we achieve the top performance on the object boundary detection task, outperforming by a large margin the current fully supervised state-of-the-art methods.

 


Figure 1. Object-specific boundaries 1a differ from generic boundaries (such as the ones detected in 1d). The proposed weakly supervised approach drives boundary detection towards the objects of interest. Example results in 1e and 1f. Red/green indicate false/true positive pixels, grey is missing recall. All methods shown at 50% recall.

Data

 

For further information or data, please contact Anna Khoreva <khoreva at mpi-inf.mpg.de>.

References

[Khoreva et al., 2016] Weakly Supervised Object Boundaries, A. Khoreva, R. Benenson, M. Omran,  M. Hein and B. Schiele, Computer Vision and Pattern Recognition (CVPR), June, (2016), (spotlight)

 

@inproceedings{khoreva16cvpr,
title={Weakly Supervised Object Boundaries
},
author={A. Khoreva and R. Benenson and M. Omran and M. Hein and B. Schiele},
booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
year={2016}}