View and Rate-Invariant Human Action Recognition
View and rate-invariant human action recognition is the recognition of actions independent of the camera viewpoint, action speed, and frame rate of capture of the video.
At its essence, human action is the movement of the body through a sequence of poses. The visual appearance of a given action in a video sequence depends upon several classes of variables: (1) the geometry of the person performing the action, (2) the style of the action being performed, (3) the clothing worn by the person, (4) the camera viewpoint, and (5) the time taken, not only for the entire action but also for each individual pose transition to complete. A generally applicable human action recognition system needs the ability to classify an action from its visual appearance regardless of the values of the above classes of variables. In other words, the system needs to be subject invariant, style invariant, clothing invariant, view...
- 1.. Bobick A, Tanawongsuwan R (2003) Performance analysis of time-distance gait parameters under different speeds. In: Proceedings of the 4th international conference on IEEE conference on computer vision and pattern recognition (CVPR), MadisonGoogle Scholar
- 2.. Jiang H (2011) Human pose estimation using consistent max-covering. Pattern Anal Mach Intell IEEE Trans PP(99):1Google Scholar
- 4.. Maurel P, Sapiro G Dynamic shapes average. www.ima.umn.edu/preprints/may2003/1924.pdf
- 5.Ogale A, Karapurkar A, Aloimonos Y (2007) View-invariant modeling and recognition of human actions using grammars. In: Vidal R, Heyden A, Ma Yi (eds) Dynamical vision. Volume 4358 of lecture notes in computer science. Springer, Berlin/Heidelberg, pp 115–126Google Scholar
- 8.. Rabiner L, Juang B (1993) Fundamentals of speech recognition. Prentice Hall, Englewood Cliffs. http://www.amazon.com/Fundamentals-Speech-Recognition-Lawrence-Rabiner/dp/0130151572
- 10.. Ratanamahatana CA, Keogh E (2004) Making time-series classification more accurate using learned constraints. In: Proceedings of the SIAM international conference on data mining, Orlando pp 11–22Google Scholar
- 14.. Souvenir R, Parrigan K (2009) Viewpoint manifolds for action recognition. EURASIP J Image Video ProcessGoogle Scholar
- 15.. Veeraraghavan A, Chellappa R, Roy-Chowdhury AK (2006) The function space of an activity. In: Proceedings of the 2006 IEEE computer society conference on computer vision and pattern recognition (CVPR), New York, vol 1, pp 959–968Google Scholar
- 18.. Weinland D, Boyer E, Ronfard R (2007) Action recognition from arbitrary views using 3D exemplars. In: Proceeding of the IEEE 11th international conference on computer vision, ICCV 2007, Rio de Janeiro, 14–21 Oct 2007, pp 1, 7Google Scholar