Papers › Transformer Based Multi-Grained Features for Unsupervised Person Re-Identification

Transformer Based Multi-Grained Features for Unsupervised Person Re-Identification

22 Nov 2022arXiv:2211.12280archive 2025-07-28

Jiachen Li, Menglin Wang, Xiaojin Gong

Multi-grained features extracted from convolutional neural networks (CNNs) have demonstrated their strong discrimination ability in supervised person re-identification (Re-ID) tasks. Inspired by them, this work investigates the way of extracting multi-grained features from a pure transformer network to address the unsupervised Re-ID problem that is label-free but much more challenging. To this end, we build a dual-branch network architecture based upon a modified Vision Transformer (ViT). The local tokens output in each branch are reshaped and then uniformly partitioned into multiple stripes to generate part-level features, while the global tokens of two branches are averaged to produce a global feature. Further, based upon offline-online associated camera-aware proxies (O2CAP) that is a top-performing unsupervised Re-ID method, we define offline and online contrastive learning losses with respect to both global and part-level features to conduct unsupervised learning. Extensive experiments on three person Re-ID datasets show that the proposed method outperforms state-of-the-art unsupervised methods by a considerable margin, greatly mitigating the gap to supervised counterparts. Code will be available soon at https://github.com/RikoLi/WACV23-workshop-TMGF.

Paper PDF Code

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

rikoli/wacv23-workshop-tmgf officialmentioned in paperpytorch 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 Person Re-Identification Unsupervised Person Re-Identification

Results from the paper archive 2025-07-28

Task	Dataset	Model	Metric	Value	Rank at snapshot	Leaderboard	Report
Unsupervised Person Re-Identification	DukeMTMC-reID	TMGF	MAP	76.8	#1 of 13	Archive leaderboard	report
Unsupervised Person Re-Identification	DukeMTMC-reID	TMGF	Rank-1	86.7	#1 of 13	Archive leaderboard	report
Unsupervised Person Re-Identification	DukeMTMC-reID	TMGF	Rank-10	94.1	#1 of 13	Archive leaderboard	report
Unsupervised Person Re-Identification	DukeMTMC-reID	TMGF	Rank-5	92.9	#1 of 13	Archive leaderboard	report
Unsupervised Person Re-Identification	MSMT17	TMGF	Rank-1	83.3	#2 of 12	Archive leaderboard	report
Unsupervised Person Re-Identification	MSMT17	TMGF	Rank-10	92.1	#2 of 12	Archive leaderboard	report
Unsupervised Person Re-Identification	MSMT17	TMGF	Rank-5	90.2	#2 of 12	Archive leaderboard	report
Unsupervised Person Re-Identification	MSMT17	TMGF	mAP	58.2	#2 of 12	Archive leaderboard	report
Unsupervised Person Re-Identification	Market-1501	TMGF	MAP	89.5	#2 of 23	Archive leaderboard	report
Unsupervised Person Re-Identification	Market-1501	TMGF	Rank-1	95.5	#2 of 23	Archive leaderboard	report
Unsupervised Person Re-Identification	Market-1501	TMGF	Rank-10	98.7	#2 of 23	Archive leaderboard	report
Unsupervised Person Re-Identification	Market-1501	TMGF	Rank-5	98.0	#2 of 23	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 Contrastive Learning Dense Connections Dropout Label Smoothing Layer Normalization Linear Layer Multi-Head Attention Position-Wise Feed-Forward Layer Residual Connection Softmax Transformer Vision 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