Discriminative figure-centric models for joint action localization and recognition
Authors:
- 1 November 2011
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 15505499,p. 2003-2010
Abstract
In this paper we develop an algorithm for action recognition and localization in videos. The algorithm uses a figure-centric visual word representation. Different from previous approaches it does not require reliable human detection and tracking as input. Instead, the person location is treated as a latent variable that is inferred simultaneously with action recognition. A spatial model for an action is learned in a discriminative fashion under a figure-centric representation. Temporal smoothness over video sequences is also enforced. We present results on the UCF-Sports dataset, verifying the effectiveness of our model in situations where detection and tracking of individuals is challenging.Keywords
This publication has 21 references indexed in Scilit:
- A segmentation-aware object detection model with occlusion handling2011
- Learning a hierarchy of discriminative space-time neighborhood features for human action recognition2010
- Spatio-temporal relationship match: Video structure comparison for recognition of complex human activities2009
- Actions in context2009
- Object detection using a max-margin Hough transform2009
- Evaluation of local spatio-temporal features for action recognition2009
- Tracking and recognizing actions of multiple hockey players using the boosted particle filterImage and Vision Computing, 2009
- Learning realistic human actions from movies2008
- Unsupervised Learning of Human Action Categories Using Spatial-Temporal WordsInternational Journal of Computer Vision, 2008
- Space-Time Behavior Based Correlation2005