Semantics-aware visual object tracking
In this paper, we propose a semantics-aware visual object tracking method, which introduces semantics into the tracking procedure and extends the model of an object with explicit semantics prior to enhancing the robustness of three key aspects of the tracking framework, i.e., appearance model, searc...
Saved in:
Main Authors: | , , , , |
---|---|
Other Authors: | |
Format: | Article |
Language: | English |
Published: |
2019
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/107568 http://hdl.handle.net/10220/50323 http://dx.doi.org/10.1109/TCSVT.2018.2848358 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-107568 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-1075682019-12-06T22:34:19Z Semantics-aware visual object tracking Shen, Chunhua Yao, Rui Lin, Guosheng Zhang, Yanning Shi, Qinfeng School of Computer Science and Engineering Semantics Engineering::Computer science and engineering Visual Object Tracking In this paper, we propose a semantics-aware visual object tracking method, which introduces semantics into the tracking procedure and extends the model of an object with explicit semantics prior to enhancing the robustness of three key aspects of the tracking framework, i.e., appearance model, search scheme, and scale adaptation. We first present a semantic object proposal generation method for video sequences to generate high-quality category-oriented object proposals. Then, a hybrid semantics-aware tracking algorithm with semantic compatibility is proposed. This algorithm takes full advantages of globally sparse semantic object proposal prediction and locally dense prediction with a template model and semantic distractor-aware color appearance model. Furthermore, we propose to exploit semantics to localize object accurately via an energy minimization framework-based scale adaptation method, which jointly integrates dense location prior, instance-specific color, and category-specific semantic information. Extensive experiments are conducted on two widely used benchmarks, and the results demonstrate that our method achieves the state-of-the-art performance. Accepted version 2019-11-04T08:39:38Z 2019-12-06T22:34:19Z 2019-11-04T08:39:38Z 2019-12-06T22:34:19Z 2018 Journal Article Yao, R., Lin, G., Shen, C., Zhang, Y., & Shi, Q. (2019). Semantics-aware visual object tracking. IEEE Transactions on Circuits and Systems for Video Technology, 29(6), 1687-1700. doi:10.1109/TCSVT.2018.2848358 1051-8215 https://hdl.handle.net/10356/107568 http://hdl.handle.net/10220/50323 http://dx.doi.org/10.1109/TCSVT.2018.2848358 en IEEE Transactions on Circuits and Systems for Video Technology © 2018 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. The published version is available at: https://doi.org/10.1109/TCSVT.2018.2848358. 14 p. application/pdf |
institution |
Nanyang Technological University |
building |
NTU Library |
country |
Singapore |
collection |
DR-NTU |
language |
English |
topic |
Semantics Engineering::Computer science and engineering Visual Object Tracking |
spellingShingle |
Semantics Engineering::Computer science and engineering Visual Object Tracking Shen, Chunhua Yao, Rui Lin, Guosheng Zhang, Yanning Shi, Qinfeng Semantics-aware visual object tracking |
description |
In this paper, we propose a semantics-aware visual object tracking method, which introduces semantics into the tracking procedure and extends the model of an object with explicit semantics prior to enhancing the robustness of three key aspects of the tracking framework, i.e., appearance model, search scheme, and scale adaptation. We first present a semantic object proposal generation method for video sequences to generate high-quality category-oriented object proposals. Then, a hybrid semantics-aware tracking algorithm with semantic compatibility is proposed. This algorithm takes full advantages of globally sparse semantic object proposal prediction and locally dense prediction with a template model and semantic distractor-aware color appearance model. Furthermore, we propose to exploit semantics to localize object accurately via an energy minimization framework-based scale adaptation method, which jointly integrates dense location prior, instance-specific color, and category-specific semantic information. Extensive experiments are conducted on two widely used benchmarks, and the results demonstrate that our method achieves the state-of-the-art performance. |
author2 |
School of Computer Science and Engineering |
author_facet |
School of Computer Science and Engineering Shen, Chunhua Yao, Rui Lin, Guosheng Zhang, Yanning Shi, Qinfeng |
format |
Article |
author |
Shen, Chunhua Yao, Rui Lin, Guosheng Zhang, Yanning Shi, Qinfeng |
author_sort |
Shen, Chunhua |
title |
Semantics-aware visual object tracking |
title_short |
Semantics-aware visual object tracking |
title_full |
Semantics-aware visual object tracking |
title_fullStr |
Semantics-aware visual object tracking |
title_full_unstemmed |
Semantics-aware visual object tracking |
title_sort |
semantics-aware visual object tracking |
publishDate |
2019 |
url |
https://hdl.handle.net/10356/107568 http://hdl.handle.net/10220/50323 http://dx.doi.org/10.1109/TCSVT.2018.2848358 |
_version_ |
1681048550988840960 |