DocumentCode :
1600867
Title :
Incremental learning of novel activity categories from videos
Author :
Ryoo, M.S. ; Joung, Jihoon ; Choi, Sunglok ; Yu, Wonpil
Author_Institution :
Robot/Cognition Res. Dept., Electron. & Telecommun. Res. Inst., Daejeon, South Korea
fYear :
2010
Firstpage :
21
Lastpage :
26
Abstract :
We present a methodology for learning novel human activities incrementally. In many real-world scenarios (e.g. YouTube), new videos of novel activities are provided additively, and the system must incrementally adjust its activity models rather than retraining the entire system after each addition. We introduce our incremental codebook learning algorithm for an efficient mining of important visual words for human activities, and propose a method that incrementally trains activity models using them. The experimental results show that our approach successfully learns human activities from increasing number of training videos, while maintaining its recognition performance comparable to previous non-incremental systems.
Keywords :
image motion analysis; learning (artificial intelligence); video signal processing; activity category; human activities learning; incremental codebook learning algorithm; training video; Feature extraction; Hidden Markov models; Histograms; Humans; Training; Videos; Visualization;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Virtual Systems and Multimedia (VSMM), 2010 16th International Conference on
Conference_Location :
Seoul
Print_ISBN :
978-1-4244-9027-1
Electronic_ISBN :
978-1-4244-9026-4
Type :
conf
DOI :
10.1109/VSMM.2010.5665972
Filename :
5665972
Link To Document :
بازگشت