Papers › Contrastive Learning for Multi-Object Tracking with Transformers

Contrastive Learning for Multi-Object Tracking with Transformers

14 Nov 2023arXiv:2311.08043archive 2025-07-28

Pierre-François De Plaen, Nicola Marinello, Marc Proesmans, Tinne Tuytelaars, Luc van Gool

The DEtection TRansformer (DETR) opened new possibilities for object detection by modeling it as a translation task: converting image features into object-level representations. Previous works typically add expensive modules to DETR to perform Multi-Object Tracking (MOT), resulting in more complicated architectures. We instead show how DETR can be turned into a MOT model by employing an instance-level contrastive loss, a revised sampling strategy and a lightweight assignment method. Our training scheme learns object appearances while preserving detection capabilities and with little overhead. Its performance surpasses the previous state-of-the-art by +2.6 mMOTA on the challenging BDD100K dataset and is comparable to existing transformer-based methods on the MOT17 dataset.

Paper PDF Code

Code

pfdp0/ContrasTR officialmentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Contrastive Learning Multi-Object Tracking Multiple Object Tracking Object Object Detection Object Tracking object-detection

Results from the paper archive 2025-07-28

Task	Dataset	Model	Metric	Value	Rank at snapshot	Leaderboard	Report
Multi-Object Tracking	MOT17	ContrasTR	HOTA	58.9	#27 of 48	Archive leaderboard	report
Multi-Object Tracking	MOT17	ContrasTR	IDF1	71.8	#27 of 48	Archive leaderboard	report
Multi-Object Tracking	MOT17	ContrasTR	MOTA	73.7	#27 of 48	Archive leaderboard	report
Multiple Object Tracking	BDD100K test	ContrasTR	mHOTA	46.1	#1 of 5	Archive leaderboard	report
Multiple Object Tracking	BDD100K test	ContrasTR	mIDF1	56.5	#1 of 5	Archive leaderboard	report
Multiple Object Tracking	BDD100K test	ContrasTR	mMOTA	42.8	#1 of 5	Archive leaderboard	report
Multiple Object Tracking	BDD100K val	ContrasTR	AssocA	-	#5 of 9	Archive leaderboard	report
Multiple Object Tracking	BDD100K val	ContrasTR	TETA	-	#5 of 9	Archive leaderboard	report
Multiple Object Tracking	BDD100K val	ContrasTR	mIDF1	52.9	#5 of 9	Archive leaderboard	report
Multiple Object Tracking	BDD100K val	ContrasTR	mMOTA	41.7	#5 of 9	Archive leaderboard	report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Absolute Position Encodings Adam Attention BPE Convolution Dense Connections Detr Dropout Feedforward Network Label Smoothing Layer Normalization Linear Layer Multi-Head Attention Position-Wise Feed-Forward Layer Residual Connection Softmax Transformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections