Learning to Segment Under Various Forms of Weak Supervision

Jia Xu1      Alexander G. Schwing2      Raquel Urtasun2

1University of Wisconsin-Madison         2University of Toronto

 
teaser-fig.png
 

Abstract

Despite the promising performance of conventional fully supervised algorithms, semantic segmentation has remained an important, yet challenging task. Due to the limited availability of complete annotations, it is of great interest to design solutions for semantic segmentation that take into account weakly labeled data, which is readily available at a much larger scale. Contrasting the common theme to develop a different algorithm for each type of weak annotation, in this work, we propose a unified approach that incorporates various forms of weak supervision -- image level tags, bounding boxes, and partial labels -- to produce a pixel-wise labeling. We conduct a rigorous evaluation on the challenging Siftflow dataset for various weakly labeled settings, and show that our approach outperforms the state-of-the-art by 12% on per-class accuracy, while maintaining comparable per-pixel accuracy.


Publication

  • Jia Xu, Alexander G. Schwing, Raquel Urtasun. Learning to Segment Under Various Forms of Weak Supervision. In IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR), June 2015. PDF, Supplement, Poster, Bibtex.

  • Source code

    Email Jia for the link.

    Acknowledgments

    We thank NVIDIA Corporation for the donation of GPUs used in this research. This work was partially funded by NSF RI 1116584 and ONR-N00014-14-1-0232.