Multiple Object Tracking of Drone Videos by a Temporal-Association Network with Separated-Tasks Structure
The task of multi-object tracking via deep learning methods for UAV videos has become an important research direction. However, with some current multiple object tracking methods, the relationship between object detection and tracking is not well handled, and decisions on how to make good use of tem...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-08-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/16/3862 |
| _version_ | 1797408171195629568 |
|---|---|
| author | Yeneng Lin Mengmeng Wang Wenzhou Chen Wang Gao Lei Li Yong Liu |
| author_facet | Yeneng Lin Mengmeng Wang Wenzhou Chen Wang Gao Lei Li Yong Liu |
| author_sort | Yeneng Lin |
| collection | DOAJ |
| description | The task of multi-object tracking via deep learning methods for UAV videos has become an important research direction. However, with some current multiple object tracking methods, the relationship between object detection and tracking is not well handled, and decisions on how to make good use of temporal information can affect tracking performance as well. To improve the performance of multi-object tracking, this paper proposes an improved multiple object tracking model based on FairMOT. The proposed model contains a structure to separate the detection and ReID heads to decrease the influence between every function head. Additionally, we develop a temporal embedding structure to strengthen the representational ability of the model. By combing the temporal-association structure and separating different function heads, the model’s performance in object detection and tracking tasks is improved, which has been verified on the VisDrone2019 dataset. Compared with the original method, the proposed model improves MOTA by 4.9% and MOTP by 1.2% and has better tracking performance than the models such as SORT and HDHNet on the UAV video dataset. |
| first_indexed | 2024-03-09T03:54:24Z |
| format | Article |
| id | doaj.art-5cefcfd681a74a2292dbae8472981992 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T03:54:24Z |
| publishDate | 2022-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-5cefcfd681a74a2292dbae84729819922023-12-03T14:23:28ZengMDPI AGRemote Sensing2072-42922022-08-011416386210.3390/rs14163862Multiple Object Tracking of Drone Videos by a Temporal-Association Network with Separated-Tasks StructureYeneng Lin0Mengmeng Wang1Wenzhou Chen2Wang Gao3Lei Li4Yong Liu5Institute of Cyber-Systems and Control, Zhejiang University, Hangzhou 310027, ChinaInstitute of Cyber-Systems and Control, Zhejiang University, Hangzhou 310027, ChinaInstitute of Cyber-Systems and Control, Zhejiang University, Hangzhou 310027, ChinaScience and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory, Beijing 100191, ChinaScience and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory, Beijing 100191, ChinaInstitute of Cyber-Systems and Control, Zhejiang University, Hangzhou 310027, ChinaThe task of multi-object tracking via deep learning methods for UAV videos has become an important research direction. However, with some current multiple object tracking methods, the relationship between object detection and tracking is not well handled, and decisions on how to make good use of temporal information can affect tracking performance as well. To improve the performance of multi-object tracking, this paper proposes an improved multiple object tracking model based on FairMOT. The proposed model contains a structure to separate the detection and ReID heads to decrease the influence between every function head. Additionally, we develop a temporal embedding structure to strengthen the representational ability of the model. By combing the temporal-association structure and separating different function heads, the model’s performance in object detection and tracking tasks is improved, which has been verified on the VisDrone2019 dataset. Compared with the original method, the proposed model improves MOTA by 4.9% and MOTP by 1.2% and has better tracking performance than the models such as SORT and HDHNet on the UAV video dataset.https://www.mdpi.com/2072-4292/14/16/3862deep learningmulti-object trackingdata associationobject detectiontemporal informationremote sensing data |
| spellingShingle | Yeneng Lin Mengmeng Wang Wenzhou Chen Wang Gao Lei Li Yong Liu Multiple Object Tracking of Drone Videos by a Temporal-Association Network with Separated-Tasks Structure Remote Sensing deep learning multi-object tracking data association object detection temporal information remote sensing data |
| title | Multiple Object Tracking of Drone Videos by a Temporal-Association Network with Separated-Tasks Structure |
| title_full | Multiple Object Tracking of Drone Videos by a Temporal-Association Network with Separated-Tasks Structure |
| title_fullStr | Multiple Object Tracking of Drone Videos by a Temporal-Association Network with Separated-Tasks Structure |
| title_full_unstemmed | Multiple Object Tracking of Drone Videos by a Temporal-Association Network with Separated-Tasks Structure |
| title_short | Multiple Object Tracking of Drone Videos by a Temporal-Association Network with Separated-Tasks Structure |
| title_sort | multiple object tracking of drone videos by a temporal association network with separated tasks structure |
| topic | deep learning multi-object tracking data association object detection temporal information remote sensing data |
| url | https://www.mdpi.com/2072-4292/14/16/3862 |
| work_keys_str_mv | AT yenenglin multipleobjecttrackingofdronevideosbyatemporalassociationnetworkwithseparatedtasksstructure AT mengmengwang multipleobjecttrackingofdronevideosbyatemporalassociationnetworkwithseparatedtasksstructure AT wenzhouchen multipleobjecttrackingofdronevideosbyatemporalassociationnetworkwithseparatedtasksstructure AT wanggao multipleobjecttrackingofdronevideosbyatemporalassociationnetworkwithseparatedtasksstructure AT leili multipleobjecttrackingofdronevideosbyatemporalassociationnetworkwithseparatedtasksstructure AT yongliu multipleobjecttrackingofdronevideosbyatemporalassociationnetworkwithseparatedtasksstructure |