A Robust Boosting Tracker with Minimum Error Bound in a Co-Training Framework
Abstract
The varying object appearance and unlabeled data from new frames are always the challenging problem in object tracking. Recently machine learning methods are widely applied to tracking, and some online and semi-supervised algorithms are developed to handle these difficulties. In this paper, we consider tracking as a classification prob-lem and present a novel tracking method based on boost-ing in a co-training framework. The proposed tracker can be online updated and boosted with multi-view weak hy-pothesis. The most important contribution of this paper is that we find a boosting error upper bound in a co-training framework to guide the novel tracker construction. In the-ory, the proposed tracking method is proved to minimize this error bound. In experiments, the accuracy rate of fore-ground/background classification and the tracking results are both served as evaluation metrics. Experimental results show good performance of proposed novel tracker on chal-lenging sequences. 1.
Cite
Text
Liu et al. "A Robust Boosting Tracker with Minimum Error Bound in a Co-Training Framework." IEEE/CVF International Conference on Computer Vision, 2009. doi:10.1109/ICCV.2009.5459285Markdown
[Liu et al. "A Robust Boosting Tracker with Minimum Error Bound in a Co-Training Framework." IEEE/CVF International Conference on Computer Vision, 2009.](https://mlanthology.org/iccv/2009/liu2009iccv-robust/) doi:10.1109/ICCV.2009.5459285BibTeX
@inproceedings{liu2009iccv-robust,
title = {{A Robust Boosting Tracker with Minimum Error Bound in a Co-Training Framework}},
author = {Liu, Rong and Cheng, Jian and Lu, Hanqing},
booktitle = {IEEE/CVF International Conference on Computer Vision},
year = {2009},
pages = {1459-1466},
doi = {10.1109/ICCV.2009.5459285},
url = {https://mlanthology.org/iccv/2009/liu2009iccv-robust/}
}