Recognizing Human Group Behaviors with Multi-group Causalities

Author

Cong Zhang ; Xiaokang Yang ; Weiyao Lin ; Jun Zhu

Author_Institution

Shanghai Key Labs. of Digital Media Process. & Commun., Shanghai Jiaotong Univ., Shanghai, China

Volume

3

fYear

2012

fDate

4-7 Dec. 2012

Firstpage

44

Lastpage

48

Abstract

Human group behaviors are usually composed of several sub-groups. Considering the interaction between groups, this paper presents an algorithm to recognize human group behavior with multi-group causalities. It has two main contributions: (1) we introduce inter-group causality to reflect the interaction between human groups, (2) an improved coding scheme (i.e. Locality-constrained Linear Coding) is used for encoding the causality to go beyond Vector Quantization (VQ). Finally, a simple linear SVM is adopted to learn this model. Our experiment results demonstrate that inter-group causality feature and LLC methods can significantly boost behavior recognition performance.

Keywords

behavioural sciences computing; computer vision; feature extraction; image coding; image recognition; support vector machines; vector quantisation; LLC method; coding scheme; computer vision; group interaction; human group behavior recognition; inter-group causality feature; locality-constrained linear coding; multigroup causality; simple linear SVM; support vector machines; vector quantization; Human group behaviors; Inter-group causality; LLC;

fLanguage

English

Publisher

ieee

Conference_Titel

Web Intelligence and Intelligent Agent Technology (WI-IAT), 2012 IEEE/WIC/ACM International Conferences on

Conference_Location

Macau

Print_ISBN

978-1-4673-6057-9

Type

conf

DOI

10.1109/WI-IAT.2012.162

Filename

6511646