Tracklet Siamese Network with Constrained Clustering for Multiple Object Tracking


See, John Su Yang and Peng, Jinlong and Qiu, Fan and Guo, Qi and Huang, Shaoshuai and Ling, Yu Duan and Lin, Weiyao (2019) Tracklet Siamese Network with Constrained Clustering for Multiple Object Tracking. In: 33rd IEEE International Conference on Visual Communications and Image Processing, VCIP 2018, 9-12 December 2018, Taichung, Taiwan.

[img] Text
162.pdf - Published Version
Restricted to Repository staff only

Download (195kB)


Multiple object tracking (MOT) is an important yet challenging task in video understanding and analysis. Basically, MOT aims to associate detected objects into trajectories based on their temporal relationships. The occlusion among moving objects poses a major challenge towards robust modeling of these relationships. In this paper, we propose a novel Tracklet Siamese Network (TSN) for learning similarities between track-lets characterized by appearance information, achieving superior performance on two MOTChallenge benchmark datasets. Our framework constructs short tracklets from highly-related object detections by excluding inaccurate object detections. We also adopt a constrained clustering technique to piece tracklets together into long trajectories, thus recovering many missing detections caused by original detector or the detection removing in the previous step. Comparisons against state-of-the-art methods were reported while ablation studies further substantiate the viability of components in our approach.

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: Convolutional neural networks
Subjects: Q Science > QA Mathematics > QA71-90 Instruments and machines
Divisions: Faculty of Computing and Informatics (FCI)
Depositing User: Ms Suzilawati Abu Samah
Date Deposited: 04 Feb 2022 03:49
Last Modified: 04 Feb 2022 03:49


Downloads per month over past year

View ItemEdit (login required)