OcclusionTrack: Multi-Object Tracking in Dense Scenes
Source abstract
Abstract
This paper presents OcclusionTrack (OCCTrack), a robust multi-object tracker designed to address occlusion challenges in dense scenes. Occlusion remains a critical issue in multi-object tracking; despite significant advancements in current tracking methods, dense scenes and frequent occlusions continue to pose formidable challenges for existing tracking-by-detection trackers. Therefore, four key improvements are integrated into a tracking-by-detection paradigm: (1) a confidence-based Kalman filter (CBKF) that dynamically adapts measurement noise to handle partial occlusions; (2) camera motion compensation (CMC) for inter-frame alignment to stabilize predictions; (3) a depth–cascade-matching (DCM) algorithm that uses relative depth to resolve association ambiguities among overlapping objects; and (4) a CMC-detection-based trajectory Re-activate method to recover and correct tracks after complete occlusion. Despite relying solely on IoU matching, OCCTrack achieves highly competitive performance on MOT17 (HOTA 64.9, MOTA 80.9, IDF1 79.7), MOT20 (HOTA 63.2, MOTA 76.9, IDF1 77.5), and DanceTrack (HOTA 57.5, MOTA 91.4, IDF1 58.4). The primary contribution of this work lies in the cohesive integration of these modules into a unified, real-time pipeline that systematically mitigates both partial and complete occlusion effects, offering a practical and reproducible framework for complex real-world tracking scenarios.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Video Surveillance and Tracking Methods · Human Pose and Action Recognition · Robotics and Sensor-Based Localization
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
School of Automation and Information Engineering, Sichuan University of Science and Engineering, Yibin 644000, China
Intelligent Perception and Control Key Laboratory of Sichuan Province, Yibin 644000, China
References (35)
Show source-provided references
1. Bewley, A., Ge, Z., Ott, L., Ramos, F., and Upadhya, A. (2016, January 25–28). Simple Online and Realtime Tracking. Proceedings of the 2016 IEEE International Conference on Image Processing (ICIP), Phoenix, AZ, USA.10.1109/ICIP.2016.7533003 ↗
2. Wojke, N., Bewley, A., and Paulus, D. (2017, January 17–20). Simple online and realtime tracking with a deep association metric. Proceedings of the IEEE International Conference on Image Processing, Beijing, China. Available online: https://ieeexplore.ieee.org/document/8296962.10.1109/ICIP.2017.8296962 ↗
3. Zhang, Y., Sun, P., Jiang, Y., Yu, D., Weng, F., Yuan, Z., Luo, P., Liu, W., and Wang, X. (2022, January 23–27). Bytetrack: Multi-object tracking byassociating every detection box. Proceedings of the European Conference on Computer Vision, Tel Aviv, Israel.10.1007/978-3-031-20047-2_1 ↗
4. Kalman, R.E. (2001). A New Approach to Linear Filtering and Prediction Problems, Wiley Press. Available online: https://ieeexplore.ieee.org/document/5311910.
5. Aharon, N., Orfaig, R., and Bobrovsky, B.Z. (2022). BoT-SORT: Robust Associations Multi-Pedestrian Tracking. arXiv.
6. Maggiolino, G., Ahmad, A., Cao, J., and Kitani, K. (2023, January 8–11). Deep OC-SORT: Multi-pedestrian tracking by adaptive re-identification. Proceedings of the 2023 IEEE International Conference on Image Processing (ICIP), Kuala Lumpur, Malaysia. Available online: https://ieeexplore.ieee.org/document/10222576.10.1109/ICIP49359.2023.10222576 ↗
7. Sparsetrack: Multi-object tracking by performing scene decomposition based on pseudo-depth10.1109/TCSVT.2024.3524670 ↗
8. Strongsort: Make deepsort great again10.1109/TMM.2023.3240881 ↗
9. Cao, J., Pang, J., Weng, X., Khirodkar, R., and Kitani, K. (2023, January 17–24). Observation-Centric SORT: Rethinking SORT for Robust Multi-Object Tracking. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, BC, Canada.10.1109/CVPR52729.2023.00934 ↗
10. Lv, W., Huang, Y., Zhang, N., Lin, R.S., Han, M., and Zeng, D. (2024, January 16–22). Diffmot: A real-time diffusion-based multiple object tracker with non-linear prediction. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA.10.1109/CVPR52733.2024.01828 ↗
11. Hong, J., Li, Y., Yan, J., Wei, X., Xian, W., and Qin, Y. (2025). KalmanFormer: Integrating a Deep Motion Model into SORT for Video Multi-Object Tracking. Appl. Sci., 15, Available online: https://www.mdpi.com/2076-3417/15/17/9727.10.3390/app15179727 ↗
12. The opencv library
13. Shi, J. (1994, January 21–23). Good features to track. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA. Available online: https://ieeexplore.ieee.org/abstract/document/323794.
14. Random sample consensus: A paradigm for model fitting with applications to image analysis and automated cartography10.1145/358669.358692 ↗
15. Milan, A., Leal-Taixé, L., Reid, I., Roth, S., and Schindler, K. (2016). Mot16: A benchmark for multi-object tracking. arXiv.
16. Dendorfer, P., Rezatofighi, H., Milan, A., Shi, J., Cremers, D., Reid, I., Roth, S., Schindler, K., and Leal-Taixé, L. (2020). Mot20: A benchmark for multi object tracking in crowded scenes. arXiv.
17. Sun, Y., Zhang, W., Zhao, B., Li, L., and Wang, J. (2022, January 18–24). DanceTrack: Multi-Object Tracking in Uniform Appearance and Diverse Motion. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), New Orleans, LA, USA. Available online: https://ieeexplore.ieee.org/document/9879192.10.1109/CVPR52688.2022.02032 ↗
18. Evaluating multiple object tracking performance: The clear mot metrics10.1155/2008/246309 ↗
19. Hota: A higher order metric for evaluating multi-object tracking10.1007/s11263-020-01375-2 ↗
20. Ristani, E., Solera, F., Zou, R., Cucchiara, R., and Tomasi, C. (15–16, January 8–10). Performance measures and a data set for multi-target, multi-camera tracking. Proceedings of the European Conference on Computer Vision, Amsterdam, The Netherlands. Available online: https://link.springer.com/chapter/10.1007/978-3-319-48881-3_2.
21. Pang, B., Li, Y., Zhang, Y., Li, M., and Lu, C. (2020, January 13–19). Tubetk: Adopting tubes to track multi-object in a one-step training model. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA.10.1109/CVPR42600.2020.00634 ↗
22. Zhou, X., Koltun, V., and Krähenbühl, P. (2020, January 23–28). Tracking objects as points. Proceedings of the European Conference on Computer Vision, Glasgow, UK.10.1007/978-3-030-58548-8_28 ↗
23. TransCenter: Transformers with Dense Representations for Multiple-Object Tracking10.1109/TPAMI.2022.3225078 ↗
24. Fairmot: On the fairness of detection and re-identification in multiple object tracking10.1007/s11263-021-01513-4 ↗
25. Chu, P., Wang, J., You, Q., Ling, H., and Liu, Z. (2021). Transmot: Spatial-temporal graph transformer for multiple object tracking. arXiv, Available online: https://ieeexplore.ieee.org/abstract/document/10030267.
26. Stadler, D., and Beyerer, J. (2022, January 4–8). Modelling ambiguous assignments for multi-person tracking in crowds. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, Waikoloa, HI, USA. Available online: https://ieeexplore.ieee.org/document/9707576.10.1109/WACVW54805.2022.00019 ↗
27. Yang, M., Han, G., Yan, B., Zhang, W., Qi, J., Lu, H., and Wang, D. (2024, January 20–27). Hybrid-sort: Weak cues matter for online multi-object tracking. Proceedings of the AAAI Conference on Artificial Intelligence, Vancouver, BC, Canada.10.1609/aaai.v38i7.28471 ↗
28. Yi, K., Luo, K., Luo, X., Huang, J., Wu, H., Hu, R., and Hao, W. (2024, January 20–27). Ucmctrack: Multi-object tracking with uniform camera motion compensation. Proceedings of the AAAI Conference on Artificial Intelligence, Vancouver, BC, Canada.10.1609/aaai.v38i7.28493 ↗
29. Multiplex labeling graph for near-online tracking in crowded scenes10.1109/JIOT.2020.2996609 ↗
30. Sun, P., Cao, J., Jiang, Y., Zhang, R., Xie, E., Yuan, Z., Wang, C., and Luo, P. (2020). Transtrack: Multiple-object tracking with transformer. arXiv.
31. Li, W., Xiong, Y., Yang, S., Xu, M., Wang, Y., and Xia, W. (2021). Semi-TCL: Semi-supervised track contrastive representation learning. arXiv.
32. Yu, E., Li, Z., Han, S., and Wang, H. (2021). RelationTrack: Relation aware multiple object tracking with decoupled representation. arXiv, Available online: https://ieeexplore.ieee.org/document/9709649.
33. Qin, Z., Zhou, S., Wang, L., Duan, J., Hua, G., and Tang, W. (2023). MotionTrack: Learning Robust Short-term and Long-term Motions for Multi-Object Tracking. arXiv.10.1109/CVPR52729.2023.01720 ↗
34. Wu, J., Cao, J., Song, L., Wang, Y., Yang, M., and Yuan, J. (2021, January 20–25). Track to detect and segment: An online multi-object tracker. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA. Available online: https://ieeexplore.ieee.org/document/9578864.10.1109/CVPR46437.2021.01217 ↗
35. Zeng, F., Dong, B., Zhang, Y., Wang, T., Zhang, X., and Wei, Y. (2022, January 23–27). MOTR: End-to-end multiple-object tracking with transformer. Proceedings of the European Conference on Computer Vision, Tel Aviv, Israel.10.1007/978-3-031-19812-0_38 ↗