Robust object tracking with online multiple instance learning
Research output: Contribution to journal › Journal article › Research › peer-review
Standard
Robust object tracking with online multiple instance learning. / Babenko, Boris; Yang, Ming Hsuan; Belongie, Serge.
In: IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 33, No. 8, 5674053, 2011, p. 1619-1632.Research output: Contribution to journal › Journal article › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - Robust object tracking with online multiple instance learning
AU - Babenko, Boris
AU - Yang, Ming Hsuan
AU - Belongie, Serge
N1 - Funding Information: The authors would like to thank Kristin Branson, Piotr Dollár, David Ross, and the anonymous reviewers for valuable input. This research has been supported by US National Science Foundation (NSF) CAREER Grant #0448615, NSF IGERT Grant DGE-0333451, and US Office of Naval Research Grant #N00014-08-1-0638. Ming-Hsuan Yang is supported in part by a University of California Merced faculty start-up fund and a Google faculty award. Part of this work was performed while Boris Babenko and Ming-Hsuan Yang were at the Honda Research Institute, USA.
PY - 2011
Y1 - 2011
N2 - In this paper, we address the problem of tracking an object in a video given its location in the first frame and no other information. Recently, a class of tracking techniques called tracking by detection has been shown to give promising results at real-time speeds. These methods train a discriminative classifier in an online manner to separate the object from the background. This classifier bootstraps itself by using the current tracker state to extract positive and negative examples from the current frame. Slight inaccuracies in the tracker can therefore lead to incorrectly labeled training examples, which degrade the classifier and can cause drift. In this paper, we show that using Multiple Instance Learning (MIL) instead of traditional supervised learning avoids these problems and can therefore lead to a more robust tracker with fewer parameter tweaks. We propose a novel online MIL algorithm for object tracking that achieves superior results with real-time performance. We present thorough experimental results (both qualitative and quantitative) on a number of challenging video clips.
AB - In this paper, we address the problem of tracking an object in a video given its location in the first frame and no other information. Recently, a class of tracking techniques called tracking by detection has been shown to give promising results at real-time speeds. These methods train a discriminative classifier in an online manner to separate the object from the background. This classifier bootstraps itself by using the current tracker state to extract positive and negative examples from the current frame. Slight inaccuracies in the tracker can therefore lead to incorrectly labeled training examples, which degrade the classifier and can cause drift. In this paper, we show that using Multiple Instance Learning (MIL) instead of traditional supervised learning avoids these problems and can therefore lead to a more robust tracker with fewer parameter tweaks. We propose a novel online MIL algorithm for object tracking that achieves superior results with real-time performance. We present thorough experimental results (both qualitative and quantitative) on a number of challenging video clips.
KW - multiple instance learning
KW - online boosting
KW - Visual Tracking
UR - http://www.scopus.com/inward/record.url?scp=79959527478&partnerID=8YFLogxK
U2 - 10.1109/TPAMI.2010.226
DO - 10.1109/TPAMI.2010.226
M3 - Journal article
AN - SCOPUS:79959527478
VL - 33
SP - 1619
EP - 1632
JO - IEEE Transactions on Pattern Analysis and Machine Intelligence
JF - IEEE Transactions on Pattern Analysis and Machine Intelligence
SN - 0162-8828
IS - 8
M1 - 5674053
ER -
ID: 301831383