Title:	ACTION RECOGNITION USING COMBINED LOCAL FEATURES
Author(s):	Ivo Reznicek, Pavel Zemcik
ISBN:	978-972-8939-89-2
Editors:	Yingcai Xiao
Year:	2013
Edition:	Single
Keywords:	Action recognition, SVM, combination of features, space-time features.
Type:	Full Paper
First Page:	111
Last Page:	118
Language:	English
Cover:
Full Contents:	click to dowload
Paper Abstract:	This paper presents a new algorithm for recognition of actions based on local space-time features. The algorithm resulted from intensive research of classification and feature extraction and it is an extension of the earlier algorithms. The most important achievement is that it is shown that carefully selected combination of space-time features leads to a greater precision of recognition on some events compared with the state-of-the-art algorithms while it is comparable on all other events. The paper describes the algorithm, its main features and improvements, demonstrates the results achieved, and draws conclusions.
	Go Back