Model uncertainty guides visual object tracking

Model object trackers largely rely on the online learning of a discriminative classifier from potentially diverse sample frames. However, noisy or insufficient amounts of samples can deteriorate the classifiers' performance and cause tracking drift. Furthermore, alterations such as occlusion an...

Full description

Saved in:
Bibliographic Details
Main Authors: ZHOU, Lijun, LEDENT, Antoine, HU, Qintao, LIU, Ting, ZHANG, Jianlin, KLOFT, Marius
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2021
Subjects:
Online Access:https://ink.library.smu.edu.sg/sis_research/7204
https://ink.library.smu.edu.sg/context/sis_research/article/8207/viewcontent/16473_Article_Text_19967_1_2_20210518.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
id sg-smu-ink.sis_research-8207
record_format dspace
spelling sg-smu-ink.sis_research-82072022-08-04T08:47:40Z Model uncertainty guides visual object tracking ZHOU, Lijun LEDENT, Antoine HU, Qintao LIU, Ting ZHANG, Jianlin KLOFT, Marius Model object trackers largely rely on the online learning of a discriminative classifier from potentially diverse sample frames. However, noisy or insufficient amounts of samples can deteriorate the classifiers' performance and cause tracking drift. Furthermore, alterations such as occlusion and blurring can cause the target to be lost. In this paper, we make several improvements aimed at tackling uncertainty and improving robustness in object tracking. Our first and most important contribution is to propose a sampling method for the online learning of object trackers based on uncertainty adjustment: our method effectively selects representative sample frames to feed the discriminative branch of the tracker, while filtering out noise samples. Furthermore, to improve the robustness of the tracker to various challenging scenarios, we propose a novel data augmentation procedure, together with a specific improved backbone architecture. All our improvements fit together in one model, which we refer to as the Uncertainty Adjusted Tracker (UATracker), and can be trained in a joint and end-to-end fashion. Experiments on the LaSOT, UAV123, OTB100 and VOT2018 benchmarks demonstrate that our UATracker outperforms state-of-the-art real-time trackers by significant margins. 2021-02-01T08:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/7204 https://ink.library.smu.edu.sg/context/sis_research/article/8207/viewcontent/16473_Article_Text_19967_1_2_20210518.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Object Tracking Computer Vision Machine Learning. Artificial Intelligence and Robotics Numerical Analysis and Scientific Computing
institution Singapore Management University
building SMU Libraries
continent Asia
country Singapore
Singapore
content_provider SMU Libraries
collection InK@SMU
language English
topic Object Tracking
Computer Vision
Machine Learning.
Artificial Intelligence and Robotics
Numerical Analysis and Scientific Computing
spellingShingle Object Tracking
Computer Vision
Machine Learning.
Artificial Intelligence and Robotics
Numerical Analysis and Scientific Computing
ZHOU, Lijun
LEDENT, Antoine
HU, Qintao
LIU, Ting
ZHANG, Jianlin
KLOFT, Marius
Model uncertainty guides visual object tracking
description Model object trackers largely rely on the online learning of a discriminative classifier from potentially diverse sample frames. However, noisy or insufficient amounts of samples can deteriorate the classifiers' performance and cause tracking drift. Furthermore, alterations such as occlusion and blurring can cause the target to be lost. In this paper, we make several improvements aimed at tackling uncertainty and improving robustness in object tracking. Our first and most important contribution is to propose a sampling method for the online learning of object trackers based on uncertainty adjustment: our method effectively selects representative sample frames to feed the discriminative branch of the tracker, while filtering out noise samples. Furthermore, to improve the robustness of the tracker to various challenging scenarios, we propose a novel data augmentation procedure, together with a specific improved backbone architecture. All our improvements fit together in one model, which we refer to as the Uncertainty Adjusted Tracker (UATracker), and can be trained in a joint and end-to-end fashion. Experiments on the LaSOT, UAV123, OTB100 and VOT2018 benchmarks demonstrate that our UATracker outperforms state-of-the-art real-time trackers by significant margins.
format text
author ZHOU, Lijun
LEDENT, Antoine
HU, Qintao
LIU, Ting
ZHANG, Jianlin
KLOFT, Marius
author_facet ZHOU, Lijun
LEDENT, Antoine
HU, Qintao
LIU, Ting
ZHANG, Jianlin
KLOFT, Marius
author_sort ZHOU, Lijun
title Model uncertainty guides visual object tracking
title_short Model uncertainty guides visual object tracking
title_full Model uncertainty guides visual object tracking
title_fullStr Model uncertainty guides visual object tracking
title_full_unstemmed Model uncertainty guides visual object tracking
title_sort model uncertainty guides visual object tracking
publisher Institutional Knowledge at Singapore Management University
publishDate 2021
url https://ink.library.smu.edu.sg/sis_research/7204
https://ink.library.smu.edu.sg/context/sis_research/article/8207/viewcontent/16473_Article_Text_19967_1_2_20210518.pdf
_version_ 1770576269455392768