DocumentCode :
2356417
Title :
Discriminative Focus of Attention for Real-Time Object Detection in Video
Author :
Saptharishi, M. ; Lipchin, A. ; Lisin, D.
Author_Institution :
VideoIQ, Inc., Bedford, MA, USA
fYear :
2012
fDate :
17-19 Oct. 2012
Firstpage :
85
Lastpage :
90
Abstract :
We propose a novel object detection approach that combines the discriminative power of object category classifiers with a simple pixel level focus of attention mechanism. The pixel-level foreground/background detectors evolve to classify each pixel as either being part of an object of interest or noise. Unlike background subtraction algorithms, the decision of what is foreground is influenced by object level knowledge rather than it being an outlier to a background distribution. The approach outperforms many background subtraction techniques in challenging scenarios. Combined with the proposed focus of attention mechanism, a robust object classifier(capable of classifying known objects or rejecting noise) runs in real-time while processing 1920x1080 videos on an off-the-shelf DSP.
Keywords :
digital signal processing chips; image classification; object detection; real-time systems; video surveillance; DSP; background detectors; background distribution; background subtraction algorithms; digital signal processing; object category classifiers; object level knowledge; pixel level foreground detectors; realtime object detection; robust object classifier; video surveillance; Adaptation models; Detectors; Lighting; Noise; Object detection; Real-time systems; Streaming media; Object detection; machine vision; real time system; video surveillance;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Signal Processing Systems (SiPS), 2012 IEEE Workshop on
Conference_Location :
Quebec City, QC
ISSN :
2162-3562
Print_ISBN :
978-1-4673-2986-6
Type :
conf
DOI :
10.1109/SiPS.2012.35
Filename :
6363188
Link To Document :
بازگشت