Discriminative human action recognition using pairwise CSP classifiers

Poppe, Ronald and Poel, Mannes (2008) Discriminative human action recognition using pairwise CSP classifiers. In: 8th IEEE International Conference on Automatic Face and Gesture Recognition, FG 2008, 17-19 September 2008, Amsterdam, The Netherlands.

PDF
Restricted to UT campus only: Request a copy
222Kb

Abstract:	We present a discriminative approach to human action recognition. At the heart of our approach is the use of common spatial patterns (CSP), a spatial filter technique that transforms temporal feature data by using differences in variance between two classes. Such a transformation focusses on differences between classes, rather than on modelling each class individually. As a results, to distinguish between two classes, we can use simple distance metrics in the low-dimensional transformed space. The most likely class is found by pairwise evaluation of all discriminant functions. Our image representations are silhouette boundary gradients, spatially binned into cells. We achieve scores of approximately 96% on a standard action dataset, and show that reasonable results can be obtained when training on only a single subject. Future work is aimed at combining our approach with automatic human detection.
Item Type:	Conference or Workshop Item
Copyright:	© 2008 IEEE
Faculty:	Electrical Engineering, Mathematics and Computer Science (EEMCS)
Research Group:	Human Media Interaction (HMI)
Link to this item:	http://purl.utwente.nl/publications/62627
Official URL:	http://dx.doi.org/10.1109/AFGR.2008.4813447
Export this item as:	BibTeX EndNote HTML Citation Reference Manager

Show download statistics for this publication

Repository Staff Only: item control page

Metis ID: 255025