DocumentCode :
639574
Title :
Seeking the Strongest Rigid Detector
Author :
Benenson, Rodrigo ; Mathias, Mayeul ; Tuytelaars, Tinne ; Van Gool, Luc
Author_Institution :
ESAT-PSI-VISICS/IBBT, Katholieke Univ. Leuven, Leuven, Belgium
fYear :
2013
fDate :
23-28 June 2013
Firstpage :
3666
Lastpage :
3673
Abstract :
The current state of the art solutions for object detection describe each class by a set of models trained on discovered sub-classes (so called "components"), with each model itself composed of collections of interrelated parts (deformable models). These detectors build upon the now classic Histogram of Oriented Gradients+linear SVM combo. In this paper we revisit some of the core assumptions in HOG+SVM and show that by properly designing the feature pooling, feature selection, preprocessing, and training methods, it is possible to reach top quality, at least for pedestrian detections, using a single rigid component. Abstract We provide experiments for a large design space, that give insights into the design of classifiers, as well as relevant information for practitioners. Our best detector is fully feed-forward, has a single unified architecture, uses only histograms of oriented gradients and colour information in monocular static images, and improves over 23 other methods on the INRIA, ETH and Caltech-USA datasets, reducing the average miss-rate over HOG+SVM by more than 30%.
Keywords :
feature extraction; gradient methods; object detection; support vector machines; Caltech-USA datasets; ETH; HOG+SVM; INRIA; core assumptions; deformable models; feature pooling; feature selection; feedforward detector; histogram of oriented gradients+linear SVM combo; interrelated parts; object detection; pedestrian detections; strongest rigid detector; training methods; Deformable models; Detectors; Feature extraction; Histograms; Image color analysis; Materials; Training; objects detection; pedestrian detection;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Vision and Pattern Recognition (CVPR), 2013 IEEE Conference on
Conference_Location :
Portland, OR
ISSN :
1063-6919
Type :
conf
DOI :
10.1109/CVPR.2013.470
Filename :
6619314
Link To Document :
بازگشت