The Application of Self-similarity Matrix Recurrence Plot in mining the Relativity of Multi-view Behavior

DOI

10.2991/icismme-15.2015.200How to use a DOI?

Keywords

View diversity; low-level features; self-similarity matrix; and behavior recognition.

Abstract

In the recognition of human actions, there are certain difficulties caused by view diversity. This paper uses the algorithm of mining the relevance of the same behavior with different views based on self-similarity matrix. This algorithm references the thought of traditional recurrence plot analysis, the difference from it is this algorithm uses low-level features directly to calculate self-similarity matrix, avoiding the difficulty of traditional recurrence plot analysis in determining the optimum embedding dimension and time delay. This algorithm aims to mine the relevance of the same behavior with different views, and we apply it to build the action recognition model, realize the robust recognition of the same behavior with different views.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Chuanxu Wang
AU  - Yan Yue
PY  - 2015/07
DA  - 2015/07
TI  - The Application of Self-similarity Matrix Recurrence Plot in mining the Relativity of Multi-view Behavior
BT  - Proceedings of the First International Conference on Information Sciences, Machinery, Materials and Energy
PB  - Atlantis Press
SP  - 948
EP  - 951
SN  - 1951-6851
UR  - https://doi.org/10.2991/icismme-15.2015.200
DO  - 10.2991/icismme-15.2015.200
ID  - Wang2015/07
ER  -