Action Recognition with Improved Trajectories

Heng Wang 1 Cordelia Schmid 1
1 LEAR - Learning and recognition in vision
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : Recently dense trajectories were shown to be an efficient video representation for action recognition and achieved state-of-the-art results on a variety of datasets. This paper improves their performance by taking into account camera motion to correct them. To estimate camera motion, we match feature points between frames using SURF descriptors and dense optical flow, which are shown to be complementary. These matches are, then, used to robustly estimate a homography with RANSAC. Human motion is in general different from camera motion and generates inconsistent matches. To improve the estimation, a human detector is employed to remove these matches. Given the estimated camera motion, we remove trajectories consistent with it. We also use this estimation to cancel out camera motion from the optical flow. This significantly improves motion-based descriptors, such as HOF and MBH. Experimental results on four challenging action datasets (i.e., Hollywood2, HMDB51, Olympic Sports and UCF50) significantly outperform the current state of the art.
Document type :
Conference papers
ICCV 2013 - IEEE International Conference on Computer Vision, Dec 2013, Sydney, Australia. IEEE, pp.3551-3558, 2013, <10.1109/ICCV.2013.441>
Liste complète des métadonnées



https://hal.inria.fr/hal-00873267
Contributor : Thoth Team <>
Submitted on : Wednesday, October 16, 2013 - 11:40:57 AM
Last modification on : Tuesday, August 11, 2015 - 1:05:15 AM

Files

wang_iccv13.pdf
Files produced by the author(s)

Identifiers

Collections

Citation

Heng Wang, Cordelia Schmid. Action Recognition with Improved Trajectories. ICCV 2013 - IEEE International Conference on Computer Vision, Dec 2013, Sydney, Australia. IEEE, pp.3551-3558, 2013, <10.1109/ICCV.2013.441>. <hal-00873267v2>

Share

Metrics

Record views

10099

Document downloads

18284