Attention Modulated Multiple Object Tracking with Motion Enhancement and Dual Correlation
The one-shot multiple object tracking (MOT) framework has drawn more and more attention in the MOT research community due to its advantage in inference speed. However, the tracking accuracy of current one-shot approaches could lead to an inferior performance compared with their two-stage counterpart...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-02-01
|
Series: | Symmetry |
Subjects: | |
Online Access: | https://www.mdpi.com/2073-8994/13/2/266 |
_version_ | 1797414773040611328 |
---|---|
author | Yifeng Wang Zhijiang Zhang Ning Zhang Dan Zeng |
author_facet | Yifeng Wang Zhijiang Zhang Ning Zhang Dan Zeng |
author_sort | Yifeng Wang |
collection | DOAJ |
description | The one-shot multiple object tracking (MOT) framework has drawn more and more attention in the MOT research community due to its advantage in inference speed. However, the tracking accuracy of current one-shot approaches could lead to an inferior performance compared with their two-stage counterparts. The reasons are two-fold: one is that motion information is often neglected due to the single-image input. The other is that detection and re-identification (ReID) are two different tasks with different focuses. Joining detection and re-identification at the training stage could lead to a suboptimal performance. To alleviate the above limitations, we propose a one-shot network named Motion and Correlation-Multiple Object Tracking (MAC-MOT). MAC-MOT introduces a motion enhance attention module (MEA) and a dual correlation attention module (DCA). MEA performs differences on adjacent feature maps which enhances the motion-related features while suppressing irrelevant information. The DCA module focuses on decoupling the detection task and re-identification task to strike a balance and reduce the competition between these two tasks. Moreover, symmetry is a core design idea in our proposed framework which is reflected in Siamese-based deep learning backbone networks, the input of dual stream images, as well as a dual correlation attention module. Our proposed approach is evaluated on the popular multiple object tracking benchmarks MOT16 and MOT17. We demonstrate that the proposed MAC-MOT can achieve a better performance than the baseline state of the arts (SOTAs). |
first_indexed | 2024-03-09T05:38:41Z |
format | Article |
id | doaj.art-0fe66521d018415c8c1b0acb13a5abac |
institution | Directory Open Access Journal |
issn | 2073-8994 |
language | English |
last_indexed | 2024-03-09T05:38:41Z |
publishDate | 2021-02-01 |
publisher | MDPI AG |
record_format | Article |
series | Symmetry |
spelling | doaj.art-0fe66521d018415c8c1b0acb13a5abac2023-12-03T12:26:38ZengMDPI AGSymmetry2073-89942021-02-0113226610.3390/sym13020266Attention Modulated Multiple Object Tracking with Motion Enhancement and Dual CorrelationYifeng Wang0Zhijiang Zhang1Ning Zhang2Dan Zeng3Key Laboratory of Specialty Fiber Optics and Optical Access Networks, Joint International Research Laboratory of Specialty Fiber Optics and Advanced Communication, Shanghai Institute of Advanced Communication and Data Science, Shanghai University, Shanghai 200444, ChinaKey Laboratory of Specialty Fiber Optics and Optical Access Networks, Joint International Research Laboratory of Specialty Fiber Optics and Advanced Communication, Shanghai Institute of Advanced Communication and Data Science, Shanghai University, Shanghai 200444, ChinaJD Al Research, Mountain View, CA 94040, USAKey Laboratory of Specialty Fiber Optics and Optical Access Networks, Joint International Research Laboratory of Specialty Fiber Optics and Advanced Communication, Shanghai Institute of Advanced Communication and Data Science, Shanghai University, Shanghai 200444, ChinaThe one-shot multiple object tracking (MOT) framework has drawn more and more attention in the MOT research community due to its advantage in inference speed. However, the tracking accuracy of current one-shot approaches could lead to an inferior performance compared with their two-stage counterparts. The reasons are two-fold: one is that motion information is often neglected due to the single-image input. The other is that detection and re-identification (ReID) are two different tasks with different focuses. Joining detection and re-identification at the training stage could lead to a suboptimal performance. To alleviate the above limitations, we propose a one-shot network named Motion and Correlation-Multiple Object Tracking (MAC-MOT). MAC-MOT introduces a motion enhance attention module (MEA) and a dual correlation attention module (DCA). MEA performs differences on adjacent feature maps which enhances the motion-related features while suppressing irrelevant information. The DCA module focuses on decoupling the detection task and re-identification task to strike a balance and reduce the competition between these two tasks. Moreover, symmetry is a core design idea in our proposed framework which is reflected in Siamese-based deep learning backbone networks, the input of dual stream images, as well as a dual correlation attention module. Our proposed approach is evaluated on the popular multiple object tracking benchmarks MOT16 and MOT17. We demonstrate that the proposed MAC-MOT can achieve a better performance than the baseline state of the arts (SOTAs).https://www.mdpi.com/2073-8994/13/2/266multiple object trackingdeep learningattention mechanism |
spellingShingle | Yifeng Wang Zhijiang Zhang Ning Zhang Dan Zeng Attention Modulated Multiple Object Tracking with Motion Enhancement and Dual Correlation Symmetry multiple object tracking deep learning attention mechanism |
title | Attention Modulated Multiple Object Tracking with Motion Enhancement and Dual Correlation |
title_full | Attention Modulated Multiple Object Tracking with Motion Enhancement and Dual Correlation |
title_fullStr | Attention Modulated Multiple Object Tracking with Motion Enhancement and Dual Correlation |
title_full_unstemmed | Attention Modulated Multiple Object Tracking with Motion Enhancement and Dual Correlation |
title_short | Attention Modulated Multiple Object Tracking with Motion Enhancement and Dual Correlation |
title_sort | attention modulated multiple object tracking with motion enhancement and dual correlation |
topic | multiple object tracking deep learning attention mechanism |
url | https://www.mdpi.com/2073-8994/13/2/266 |
work_keys_str_mv | AT yifengwang attentionmodulatedmultipleobjecttrackingwithmotionenhancementanddualcorrelation AT zhijiangzhang attentionmodulatedmultipleobjecttrackingwithmotionenhancementanddualcorrelation AT ningzhang attentionmodulatedmultipleobjecttrackingwithmotionenhancementanddualcorrelation AT danzeng attentionmodulatedmultipleobjecttrackingwithmotionenhancementanddualcorrelation |