Human Action Recognition Based on Spatio-temporal Features

This paper studies the technique of human action recognition using spatio-temporal features. We concentrate on the motion and the shape patterns produced by different actions for action recognition. The motion patterns generated by the actions are captured by the optical flows. The Shape information is obtained by Viola-Jones features. Spatial features comprises of motion and shape information from a single frame. Spatio-temporal descriptor patterns are formed to improve the accuracy over spatial features. Adaboost learns and classifies the descriptor patterns. We report the accuracy of our system on a standard Weizmann dataset.

[1]  Yoav Freund,et al.  A Short Introduction to Boosting , 1999 .

[2]  Feng Niu,et al.  View-Invariant Human Activity Recognition Based on Shape and Motion Features , 2007, Int. J. Robotics Autom..

[3]  Stefan Carlsson,et al.  Recognizing and Tracking Human Action , 2002, ECCV.

[4]  Paul A. Viola,et al.  Robust Real-Time Face Detection , 2001, International Journal of Computer Vision.

[5]  Ronen Basri,et al.  Actions as space-time shapes , 2005, Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1.

[6]  Yasushi Yagi.  Computer Vision - ACCV 2007, 8th Asian Conference on Computer Vision, Tokyo, Japan, November 18-22, 2007, Proceedings, Part I , 2007, ACCV.

[7]  Somayeh Danafar,et al.  Action Recognition for Surveillance Applications Using Optic Flow and SVM , 2007, ACCV.

[8]  Dariu Gavrila,et al.  The Visual Analysis of Human Movement: A Survey , 1999, Comput. Vis. Image Underst..

[9]  Mubarak Shah,et al.  Action recognition using spatio-temporal regularity based features , 2008, 2008 IEEE International Conference on Acoustics, Speech and Signal Processing.

[10]  Mads Nielsen,et al.  Computer Vision — ECCV 2002 , 2002, Lecture Notes in Computer Science.

[11]  Takeo Kanade,et al.  An Iterative Image Registration Technique with an Application to Stereo Vision , 1981, IJCAI.