DocumentCode :
595171
Title :
Enhancing cross-view object classification by feature-based transfer learning
Author :
Yi Mo ; Zhaoxiang Zhang ; Yunhong Wang
Author_Institution :
State Key Lab. of Virtual Reality Technol. & Syst., Beihang Univ., Beijing, China
fYear :
2012
fDate :
11-15 Nov. 2012
Firstpage :
2218
Lastpage :
2221
Abstract :
Object classification is of vital importance to intelligent traffic surveillance. A big challenge is that shooting view changes in different scenes, which leads to sharp accuracy decrease since training and test samples do not follow the same distribution anymore. On the other hand, manual labeling training samples is time and labor consuming. We propose a feature-based transfer learning framework to gap the divergence of different domain distributions with scarce target view samples. Source view samples, following a different but relevant distribution, could be utilized to learn what a good classifier is like by structure learning. At the same time, small amount of target view samples could make a great contribution to reflect the target distribution. Experimental results indicate that our method outperforms traditional approaches when target samples are too scarce to build a strong classifier.
Keywords :
automated highways; feature extraction; image classification; learning (artificial intelligence); natural scenes; object recognition; road traffic; traffic engineering computing; video surveillance; classifier; cross-view object classification; domain distributions; feature-based transfer learning framework; intelligent traffic surveillance; source view samples; structure learning; target view samples; traffic scenes; Accuracy; Joints; Manuals; Pattern recognition; Surveillance; Training; Vectors;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Pattern Recognition (ICPR), 2012 21st International Conference on
Conference_Location :
Tsukuba
ISSN :
1051-4651
Print_ISBN :
978-1-4673-2216-4
Type :
conf
Filename :
6460604
Link To Document :
بازگشت