Abstract : We address the problem of segmenting multiple object instances in complex videos. Our method does not require manual pixel-level annotation for training, and relies instead on readily-available object detectors or visual object tracking only. Given object bounding boxes at input, we cast video segmentation as a weakly-supervised learning problem. Our proposed objective combines (a) a discrim-inative clustering term for background segmentation, (b) a spectral clustering one for grouping pixels of same object instances, and (c) linear constraints enabling instance-level segmentation. We propose a convex relaxation of this problem and solve it efficiently using the Frank-Wolfe algorithm. We report results and compare our method to several base-lines on a new video dataset for multi-instance person seg-mentation.
https://hal.inria.fr/hal-01255765
Contributor : Guillaume Seguin
<>
Submitted on : Wednesday, January 13, 2016 - 10:44:29 PM
Last modification on : Thursday, February 7, 2019 - 3:49:40 PM
Long-term archiving on : Friday, November 11, 2016 - 5:04:25 AM
Guillaume Seguin, Piotr Bojanowski, Rémi Lajugie, Ivan Laptev. Instance-level video segmentation from object tracks. CVPR 2016, IEEE, Jun 2016, Las Vegas, United States. ⟨hal-01255765⟩
