Person Re-Identification on MSMT17

Metric: mAP (higher is better)

LeaderboardDataset

Loading chart...

Results

Hide extra data

Sort:

#	Model↕	mAP▼	Extra Data	Paper	Date↕	Code
1	CLIP-ReID (with re-ranking)	86.7	No	CLIP-ReID: Exploiting Vision-Language Model for ...	2022-11-25	Code
2	SOLIDER (with re-ranking)	86.5	Yes	Beyond Appearance: a Semantic Controllable Self-...	2023-03-30	Code
3	FlipReID (with re-ranking)	81.3	No	FlipReID: Closing the Gap between Training and I...	2021-05-12	Code
4	ProNet++ (ResNet50+RK)	80	No	Rethinking Person Re-identification from a Proje...	2023-08-21	-
5	LDS (ResNet50+RK)	79.09	No	Learning to Disentangle Scenes for Person Re-ide...	2021-11-10	Code
6	SOLIDER (without re-ranking)	77.1	Yes	Beyond Appearance: a Semantic Controllable Self-...	2023-03-30	Code
7	Adaptive L2 Regularization (with re-ranking)	76.7	No	Adaptive L2 Regularization in Person Re-Identifi...	2020-07-15	Code
8	PCL-CLIP (L_pcl+L_id)	76.1	No	Prototypical Contrastive Learning-based CLIP Fin...	2023-10-26	Code
9	CLIP-ReID (without re-ranking)	75.8	No	CLIP-ReID: Exploiting Vision-Language Model for ...	2022-11-25	Code
10	TransReID-SSL (ViT-B without RK)	75	No	Self-Supervised Pre-Training for Transformer-Bas...	2021-11-23	Code
11	CA-Jaccard	74.1	No	CA-Jaccard: Camera-aware Jaccard Distance for Pe...	2023-11-17	Code
12	PCL-CLIP (L_pcl)	73.8	No	Prototypical Contrastive Learning-based CLIP Fin...	2023-10-26	Code
13	DiP (without RK)	71.8	No	DiP: Learning Discriminative Implicit Parts for ...	2022-12-24	Code
14	TransReID	69.4	No	TransReID: Transformer-based Object Re-Identific...	2021-02-08	Code
15	Unsupervised Pre-training (ResNet101+MGN)	68.8	Yes	Unsupervised Pre-training for Person Re-identifi...	2020-12-07	Code
16	Weakly Supervised Pre-training (ResNet50+MGN)	68	No	Large-Scale Pre-training for Person Re-identific...	2022-03-30	Code
17	FlipReID (without re-ranking)	68	No	FlipReID: Closing the Gap between Training and I...	2021-05-12	Code
18	CLIP-ReID Baseline + UFFM +AMC	67.6	No	Enhancing person re-identification via Uncertain...	2024-05-02	Code
19	Deep Miner (w/o ReRank)	67.3	No	Deep Miner: A Deep and Multi-branch Network whic...	2021-02-18	Code
20	UniHCP (finetune)	67.3	No	UniHCP: A Unified Model for Human-Centric Percep...	2023-03-06	Code
21	RGT&RGPR(without RK)	65.9	Yes	Eliminate Deviation with Deviation for Data Augm...	2021-01-21	Code
22	ProNet++	65.5	No	Rethinking Person Re-identification from a Proje...	2023-08-21	-
23	PCL-CLIP (O2CAP)	65.5	No	Prototypical Contrastive Learning-based CLIP Fin...	2023-10-26	Code
24	GiT	64.8	No	GiT: Graph Interactive Transformer for Vehicle R...	2021-07-12	-
25	CAL(ResNet50)	64	No	Counterfactual Attention Learning for Fine-Grain...	2021-08-19	Code
26	ReMix	63.9	Yes	ReMix: Training Generalized Person Re-identifica...	2024-10-29	-
27	APNet-C(ResNet50)	63.5	No	Person Re-identification via Attention Pyramid	2021-08-11	Code
28	MPN (without re-ranking)	62.7	No	Multi-task Learning with Coarse Priors for Robus...	2020-03-18	Code
29	BoT+UFFM+AMC	62.3	No	Enhancing person re-identification via Uncertain...	2024-05-02	Code
30	Adaptive L2 Regularization (without re-ranking)	62.2	No	Adaptive L2 Regularization in Person Re-Identifi...	2020-07-15	Code
31	CACENET (ResNet50 w/o RR)	62	No	Devil's in the Details: Aligning Visual Clues fo...	2020-09-11	Code
32	ABD-Net (ResNet-50)	60.8	No	ABD-Net: Attentive but Diverse Person Re-Identif...	2019-08-03	Code
33	MSINet (2.3M w/o RK)	59.6	No	MSINet: Twins Contrastive Search of Multi-Scale ...	2023-03-13	Code
34	TMGF	58.2	No	Transformer Based Multi-Grained Features for Uns...	2022-11-22	Code
35	PCL-CLIP (CC)	56.4	No	Prototypical Contrastive Learning-based CLIP Fin...	2023-10-26	Code
36	PCL-CLIP (CAP)	53.6	No	Prototypical Contrastive Learning-based CLIP Fin...	2023-10-26	Code
37	OSNet	52.9	No	Omni-Scale Feature Learning for Person Re-Identi...	2019-05-02	Code
38	DG-Net	52.3	No	Joint Discriminative and Generative Learning for...	2019-04-15	Code
39	MGN + CircleLoss(ours)	52.1	No	Circle Loss: A Unified Perspective of Pair Simil...	2020-02-25	Code
40	TransReID-SSL (ViTi-S)	50.6	Yes	Self-Supervised Pre-Training for Transformer-Bas...	2021-11-23	Code
41	ResNet50 + CircleLoss(ours)	50.2	No	Circle Loss: A Unified Perspective of Pair Simil...	2020-02-25	Code
42	OSNet-IAP 1.0x	48.66	No	Building Computationally Efficient and Well-Gene...	2020-03-17	Code
43	CA-Jaccard	44.3	No	CA-Jaccard: Camera-aware Jaccard Distance for Pe...	2023-11-17	Code
44	CBN	42.9	No	Rethinking the Distribution Gap of Person Re-ide...	2020-01-23	Code
45	PPLR	42.2	No	Part-based Pseudo Label Refinement for Unsupervi...	2022-03-28	Code
46	TransReID-SSL (ViT-S)	40.9	Yes	Self-Supervised Pre-Training for Transformer-Bas...	2021-11-23	Code
47	VAL-PAT	38.9	No	Learning Transferable Pedestrian Representation ...	2023-04-12	Code
48	Cluster Contrast	33	No	Cluster Contrast for Unsupervised Person Re-Iden...	2021-03-22	Code
49	DLCE	31.58	No	A Discriminatively Learned CNN Embedding for Per...	2016-11-17	Code
50	Group Sampling	24.6	No	Rethinking Sampling Strategies for Unsupervised ...	2021-07-07	Code
51	OGNet	23.01	No	Parameter-Efficient Person Re-identification in ...	2020-06-08	Code
52	CACL	23	No	Cluster-guided Asymmetric Contrastive Learning f...	2021-06-15	Code
53	UTAL	13.1	No	Unsupervised Tracklet Person Re-Identification	2019-03-01	Code
54	TAUDL	12.5	No	Unsupervised Person Re-identification by Deep Le...	2018-09-08	-

#1CLIP-ReID (with re-ranking)SOTA
86.7
mAP· 2022-11-25
CLIP-ReID: Exploiting Vision-Language Model for Image Re-Identification without Concrete Text Labels Code
#2SOLIDER (with re-ranking)
86.5
mAP· Extra Data· 2023-03-30
Beyond Appearance: a Semantic Controllable Self-Supervised Learning Framework for Human-Centric Visual Tasks Code
#3FlipReID (with re-ranking)SOTA
81.3
mAP· 2021-05-12
FlipReID: Closing the Gap between Training and Inference in Person Re-Identification Code
#4ProNet++ (ResNet50+RK)
80
mAP· 2023-08-21
Rethinking Person Re-identification from a Projection-on-Prototypes Perspective
#5LDS (ResNet50+RK)
79.09
mAP· 2021-11-10
Learning to Disentangle Scenes for Person Re-identification Code
#6SOLIDER (without re-ranking)
77.1
mAP· Extra Data· 2023-03-30
Beyond Appearance: a Semantic Controllable Self-Supervised Learning Framework for Human-Centric Visual Tasks Code
#7Adaptive L2 Regularization (with re-ranking)SOTA
76.7
mAP· 2020-07-15
Adaptive L2 Regularization in Person Re-Identification Code
#8PCL-CLIP (L_pcl+L_id)
76.1
mAP· 2023-10-26
Prototypical Contrastive Learning-based CLIP Fine-tuning for Object Re-identification Code
#9CLIP-ReID (without re-ranking)
75.8
mAP· 2022-11-25
CLIP-ReID: Exploiting Vision-Language Model for Image Re-Identification without Concrete Text Labels Code
#10TransReID-SSL (ViT-B without RK)
75
mAP· 2021-11-23
Self-Supervised Pre-Training for Transformer-Based Person Re-Identification Code
#11CA-Jaccard
74.1
mAP· 2023-11-17
CA-Jaccard: Camera-aware Jaccard Distance for Person Re-identification Code
#12PCL-CLIP (L_pcl)
73.8
mAP· 2023-10-26
Prototypical Contrastive Learning-based CLIP Fine-tuning for Object Re-identification Code
#13DiP (without RK)
71.8
mAP· 2022-12-24
DiP: Learning Discriminative Implicit Parts for Person Re-Identification Code
#14TransReID
69.4
mAP· 2021-02-08
TransReID: Transformer-based Object Re-Identification Code
#15Unsupervised Pre-training (ResNet101+MGN)
68.8
mAP· Extra Data· 2020-12-07
Unsupervised Pre-training for Person Re-identification Code
#16Weakly Supervised Pre-training (ResNet50+MGN)
68
mAP· 2022-03-30
Large-Scale Pre-training for Person Re-identification with Noisy Labels Code
#17FlipReID (without re-ranking)
68
mAP· 2021-05-12
FlipReID: Closing the Gap between Training and Inference in Person Re-Identification Code
#18CLIP-ReID Baseline + UFFM +AMC
67.6
mAP· 2024-05-02
Enhancing person re-identification via Uncertainty Feature Fusion Method and Auto-weighted Measure Combination Code
#19Deep Miner (w/o ReRank)
67.3
mAP· 2021-02-18
Deep Miner: A Deep and Multi-branch Network which Mines Rich and Diverse Features for Person Re-identification Code
#20UniHCP (finetune)
67.3
mAP· 2023-03-06
UniHCP: A Unified Model for Human-Centric Perceptions Code
#21RGT&RGPR(without RK)
65.9
mAP· Extra Data· 2021-01-21
Eliminate Deviation with Deviation for Data Augmentation and a General Multi-modal Data Learning Method Code
#22ProNet++
65.5
mAP· 2023-08-21
Rethinking Person Re-identification from a Projection-on-Prototypes Perspective
#23PCL-CLIP (O2CAP)
65.5
mAP· 2023-10-26
Prototypical Contrastive Learning-based CLIP Fine-tuning for Object Re-identification Code
#24GiT
64.8
mAP· 2021-07-12
GiT: Graph Interactive Transformer for Vehicle Re-identification
#25CAL(ResNet50)
64
mAP· 2021-08-19
Counterfactual Attention Learning for Fine-Grained Visual Categorization and Re-identification Code
#26ReMix
63.9
mAP· Extra Data· 2024-10-29
ReMix: Training Generalized Person Re-identification on a Mixture of Data
#27APNet-C(ResNet50)
63.5
mAP· 2021-08-11
Person Re-identification via Attention Pyramid Code
#28MPN (without re-ranking)SOTA
62.7
mAP· 2020-03-18
Multi-task Learning with Coarse Priors for Robust Part-aware Person Re-identification Code
#29BoT+UFFM+AMC
62.3
mAP· 2024-05-02
Enhancing person re-identification via Uncertainty Feature Fusion Method and Auto-weighted Measure Combination Code
#30Adaptive L2 Regularization (without re-ranking)
62.2
mAP· 2020-07-15
Adaptive L2 Regularization in Person Re-Identification Code
#31CACENET (ResNet50 w/o RR)
62
mAP· 2020-09-11
Devil's in the Details: Aligning Visual Clues for Conditional Embedding in Person Re-Identification Code
#32ABD-Net (ResNet-50)SOTA
60.8
mAP· 2019-08-03
ABD-Net: Attentive but Diverse Person Re-Identification Code
#33MSINet (2.3M w/o RK)
59.6
mAP· 2023-03-13
MSINet: Twins Contrastive Search of Multi-Scale Interaction for Object ReID Code
#34TMGF
58.2
mAP· 2022-11-22
Transformer Based Multi-Grained Features for Unsupervised Person Re-Identification Code
#35PCL-CLIP (CC)
56.4
mAP· 2023-10-26
Prototypical Contrastive Learning-based CLIP Fine-tuning for Object Re-identification Code
#36PCL-CLIP (CAP)
53.6
mAP· 2023-10-26
Prototypical Contrastive Learning-based CLIP Fine-tuning for Object Re-identification Code
#37OSNetSOTA
52.9
mAP· 2019-05-02
Omni-Scale Feature Learning for Person Re-Identification Code
#38DG-NetSOTA
52.3
mAP· 2019-04-15
Joint Discriminative and Generative Learning for Person Re-identification Code
#39MGN + CircleLoss(ours)
52.1
mAP· 2020-02-25
Circle Loss: A Unified Perspective of Pair Similarity Optimization Code
#40TransReID-SSL (ViTi-S)
50.6
mAP· Extra Data· 2021-11-23
Self-Supervised Pre-Training for Transformer-Based Person Re-Identification Code
#41ResNet50 + CircleLoss(ours)
50.2
mAP· 2020-02-25
Circle Loss: A Unified Perspective of Pair Similarity Optimization Code
#42OSNet-IAP 1.0x
48.66
mAP· 2020-03-17
Building Computationally Efficient and Well-Generalizing Person Re-Identification Models with Metric Learning Code
#43CA-Jaccard
44.3
mAP· 2023-11-17
CA-Jaccard: Camera-aware Jaccard Distance for Person Re-identification Code
#44CBN
42.9
mAP· 2020-01-23
Rethinking the Distribution Gap of Person Re-identification with Camera-based Batch Normalization Code
#45PPLR
42.2
mAP· 2022-03-28
Part-based Pseudo Label Refinement for Unsupervised Person Re-identification Code
#46TransReID-SSL (ViT-S)
40.9
mAP· Extra Data· 2021-11-23
Self-Supervised Pre-Training for Transformer-Based Person Re-Identification Code
#47VAL-PAT
38.9
mAP· 2023-04-12
Learning Transferable Pedestrian Representation from Multimodal Information Supervision Code
#48Cluster Contrast
33
mAP· 2021-03-22
Cluster Contrast for Unsupervised Person Re-Identification Code
#49DLCESOTA
31.58
mAP· 2016-11-17
A Discriminatively Learned CNN Embedding for Person Re-identification Code
#50Group Sampling
24.6
mAP· 2021-07-07
Rethinking Sampling Strategies for Unsupervised Person Re-identification Code
#51OGNet
23.01
mAP· 2020-06-08
Parameter-Efficient Person Re-identification in the 3D Space Code
#52CACL
23
mAP· 2021-06-15
Cluster-guided Asymmetric Contrastive Learning for Unsupervised Person Re-Identification Code
#53UTAL
13.1
mAP· 2019-03-01
Unsupervised Tracklet Person Re-Identification Code
#54TAUDL
12.5
mAP· 2018-09-08
Unsupervised Person Re-identification by Deep Learning Tracklet Association