ABSTRACT
Modern multi-object tracking (MOT) benefited from recent advances in deep neural network and large video datasets. However, there are still some challenges impeding further improvement of the tracking performance, including complex background, fast motion and occlusion scenes. In this paper, we propose a new framework which employs motion information with optical flow, enable directly distinguishing the foreground and background regions. The proposed end-to-end network consists of two branches to separately model the spatial feature representations and optical flow motion patterns. We propose different fusion mechanism by combining the motion clues and appearance information. The results on MOT17 dataset show that our method is an effective mechanism in modeling temporal-spatial information.
- Tak-Wai Hui, Xiaoou Tang, Chen Change Loy. 2018.LiteFlowNet: A Lightweight Convolutional Neural Network for Optical Flow Estimation.In Computer Vision & Pattern Recognition.Google Scholar
- J. Peng, C. Wang, F. Wan, Y. Wu, Y. Wang, Y. Tai, C. Wang, J. Li, F. Huang, and Y. Fu.2020. Chained-tracker: Chaining paired attentive regression results for end-to-end joint multiple-object detection and tracking. ArXiv preprint arXiv:2007.14557, 2020Google Scholar
- A. Bewley, Z. Ge, L. Ott, F. Ramos, and B. Upcroft. 2016. Simple online and realtime tracking. In ICIP. IEEE, pp. 3464–3468.Google Scholar
- Wojke, N., Bewley, A., Paulus, D. 2017. Simple online and realtime tracking with a deep association metric. In: 2017 IEEE international conference on image processing (ICIP). pp. 3645–3649.Google Scholar
- Zhou, X., Koltun, V., Kr¨ahenb¨uhl, P.2020. Tracking objects as points. In European Conference on Computer Vision. pp. 474–490. Springer.Google Scholar
- Zhang, Y., Wang, C., Wang, X., Zeng, W., Liu, W.2021.Fairmot: On the fairness of detection and re-identification in multiple object tracking. International Journal of Computer Vision 129(11), 3069–3087Google ScholarDigital Library
- J. Peng, C. Wang, F. Wan, Y. Wu, Y. Wang, Y. Tai, C. Wang, J. Li, F. Huang, and Y. Fu.2020. Chained-tracker: Chaining paired attentive regression results for end-to-end joint multiple-object detection and tracking. arXiv preprint arXiv:2007.14557.Google Scholar
- B. Pang, Y. Li, Y. Zhang, M. Li, and C. Lu. 2020. 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, pp. 6308–6318.Google Scholar
- Xu, Y., Ban, Y., Delorme, G., Gan, C., Rus, D., Alameda-Pineda, X.2021. Transcenter: Transformers with dense queries for multiple-object tracking. In arXiv preprint arXiv:2103.15145Google Scholar
- G. Welch, G. Bishop 1995. An introduction to the kalman filter.Google Scholar
- E. Ilg, N. Mayer, T. Saikia, M. Keuper, A. Dosovitskiy, and T. Brox.2017. FlowNet2.0: Evolution of optical flow estimation with deep networks. CVPR, pages 2462–2470Google Scholar
- Saad A. Yaseen and Sreela Sasi. 2014. Robust Algorithm for Object Detection and Tracking in a Dynamic Scene. In Journal of Image and Graphics, Vol. 2, No. 1, pp. 41- 45, June 2014. doi: 10.12720/joig.2.1.41-45Google ScholarCross Ref
- K. He, X. Zhang, S. Ren, and J. Sun.2016. Deep residual learning for image recognition. In CVPR. pp. 770–778.Google Scholar
- S. Shao, Z. Zhao, B. Li, T. Xiao, G. Yu, X. Zhang, and J. Sun.2018. Crowdhuman: A benchmark for detecting human in a crowd.In arXiv preprint arXiv:1805.00123Google Scholar
- P. Dollar, C. Wojek, B. Schiele, and P. Perona.2009. Pedestrian detection: A ´ benchmark. In CVPR. IEEE, pp. 304–311.Google Scholar
- K. Bernardin and R. Stiefelhagen, 2008 Evaluating multiple object tracking performance: the clear mot metrics.EURASIP Journal on Image and Video Processing, vol. 2008, pp. 1–10Google Scholar
- Karthik Dinesh and Sumana Gupta. 2014. Video Stabilization, Camera Motion Pattern Recognition and Motion Tracking Using Spatiotemporal Regularity Flow.In Journal of Image and Graphics, Vol. 2, No. 1, pp. 33-40, June 2014. doi: 10.12720/joig.2.1.33-40Google ScholarCross Ref
Index Terms
MMOT: Motion-Aware Multi-Object Tracking with Optical Flow
Recommendations
Robust object tracking via multi-cue fusion
A long-term object tracking method based on calibrated binocular cameras by fusing information of the two channels and binocular geometry constraints is proposed.The stereo filter which is built based on the epipolar geometry of the binocular cameras is ...
Improved Simultaneous Computation of Motion Detection and Optical Flow for Object Tracking
DICTA '09: Proceedings of the 2009 Digital Image Computing: Techniques and ApplicationsObject tracking systems require accurate segmentation of the objects from the background for effective tracking. Motion segmentation or optical flow can be used to segment incoming images. Whilst optical flow allows multiple moving targets to be ...
Motion Field and Optical Flow: Qualitative Properties
It is shown that the motion field the 2-D vector field which is the perspective projection on the image plane of the 3-D velocity field of a moving scene, and the optical flow, defined as the estimate of the motion field which can be derived from the ...
Comments