DMCA
Online Boosting and Vision (2006)
Cached
Download Links
| Citations: | 246 - 41 self |
BibTeX
@MISC{Grabner06onlineboosting,
author = {Helmut Grabner and Horst Bischof},
title = {Online Boosting and Vision},
year = {2006}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Boosting has become very popular in computer vision, showing impressive performance in detection and recognition tasks. Mainly off-line training methods have been used, which implies that all training data has to be a priori given; training and usage of the classifier are separate steps. Training the classifier on-line and incrementally as new data becomes available has several advantages and opens new areas of application for boosting in computer vision. In this paper we propose a novel on-line AdaBoost feature selection method. In conjunction with efficient feature extraction methods the method is real time capable. We demonstrate the multifariousness of the method on such diverse tasks as learning complex background models, visual tracking and object detection. All approaches benefit significantly by the on-line training. 1.
Keyphrases
online boosting computer vision on-line training training data new data becomes separate step complex background model recognition task efficient feature extraction visual tracking diverse task object detection off-line training method several advantage new area real time impressive performance
