Towards Distraction-Robust Active Visual Tracking

Fangwei Zhong, Peng Sun, Wenhan Luo, Tingyun Yan, Yizhou Wang
Proceedings of the 38th International Conference on Machine Learning, PMLR 139:12782-12792, 2021.

Abstract

In active visual tracking, it is notoriously difficult when distracting objects appear, as distractors often mislead the tracker by occluding the target or bringing a confusing appearance. To address this issue, we propose a mixed cooperative-competitive multi-agent game, where a target and multiple distractors form a collaborative team to play against a tracker and make it fail to follow. Through learning in our game, diverse distracting behaviors of the distractors naturally emerge, thereby exposing the tracker’s weakness, which helps enhance the distraction-robustness of the tracker. For effective learning, we then present a bunch of practical methods, including a reward function for distractors, a cross-modal teacher-student learning strategy, and a recurrent attention mechanism for the tracker. The experimental results show that our tracker performs desired distraction-robust active visual tracking and can be well generalized to unseen environments. We also show that the multi-agent game can be used to adversarially test the robustness of trackers.

Cite this Paper

BibTeX
@InProceedings{pmlr-v139-zhong21b, title = {Towards Distraction-Robust Active Visual Tracking}, author = {Zhong, Fangwei and Sun, Peng and Luo, Wenhan and Yan, Tingyun and Wang, Yizhou}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {12782--12792}, year = {2021}, editor = {Meila, Marina and Zhang, Tong}, volume = {139}, series = {Proceedings of Machine Learning Research}, month = {18--24 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v139/zhong21b/zhong21b.pdf}, url = {https://proceedings.mlr.press/v139/zhong21b.html}, abstract = {In active visual tracking, it is notoriously difficult when distracting objects appear, as distractors often mislead the tracker by occluding the target or bringing a confusing appearance. To address this issue, we propose a mixed cooperative-competitive multi-agent game, where a target and multiple distractors form a collaborative team to play against a tracker and make it fail to follow. Through learning in our game, diverse distracting behaviors of the distractors naturally emerge, thereby exposing the tracker’s weakness, which helps enhance the distraction-robustness of the tracker. For effective learning, we then present a bunch of practical methods, including a reward function for distractors, a cross-modal teacher-student learning strategy, and a recurrent attention mechanism for the tracker. The experimental results show that our tracker performs desired distraction-robust active visual tracking and can be well generalized to unseen environments. We also show that the multi-agent game can be used to adversarially test the robustness of trackers.} }
Endnote
%0 Conference Paper %T Towards Distraction-Robust Active Visual Tracking %A Fangwei Zhong %A Peng Sun %A Wenhan Luo %A Tingyun Yan %A Yizhou Wang %B Proceedings of the 38th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Marina Meila %E Tong Zhang %F pmlr-v139-zhong21b %I PMLR %P 12782--12792 %U https://proceedings.mlr.press/v139/zhong21b.html %V 139 %X In active visual tracking, it is notoriously difficult when distracting objects appear, as distractors often mislead the tracker by occluding the target or bringing a confusing appearance. To address this issue, we propose a mixed cooperative-competitive multi-agent game, where a target and multiple distractors form a collaborative team to play against a tracker and make it fail to follow. Through learning in our game, diverse distracting behaviors of the distractors naturally emerge, thereby exposing the tracker’s weakness, which helps enhance the distraction-robustness of the tracker. For effective learning, we then present a bunch of practical methods, including a reward function for distractors, a cross-modal teacher-student learning strategy, and a recurrent attention mechanism for the tracker. The experimental results show that our tracker performs desired distraction-robust active visual tracking and can be well generalized to unseen environments. We also show that the multi-agent game can be used to adversarially test the robustness of trackers.
APA
Zhong, F., Sun, P., Luo, W., Yan, T. & Wang, Y.. (2021). Towards Distraction-Robust Active Visual Tracking. Proceedings of the 38th International Conference on Machine Learning, in Proceedings of Machine Learning Research 139:12782-12792 Available from https://proceedings.mlr.press/v139/zhong21b.html.

Related Material