Visual Place Recognition on Pittsburgh-30k-test

Metric: Recall@1 (higher is better)

LeaderboardDataset

Loading chart...

Results

Hide extra data

Sort:

#	Model↕	Recall@1▼	Extra Data	Paper	Date↕	Code
1	Pair-VPR-p	95.4	No	Pair-VPR: Place-Aware Pre-training and Contrasti...	2024-10-09	Code
2	Pair-VPR-s	94.7	No	Pair-VPR: Place-Aware Pre-training and Contrasti...	2024-10-09	Code
3	FoL	94.5	No	Focus on Local: Finding Reliable Discriminative ...	2025-04-14	Code
4	QAA-DINOv2-B-8192	94.4	Yes	Query-Based Adaptive Aggregation for Multi-Datas...	2025-07-04	-
5	EffoVPR	93.9	No	EffoVPR: Effective Foundation Model Utilization ...	2024-05-28	-
6	FoL-global	93.9	No	Focus on Local: Finding Reliable Discriminative ...	2025-04-14	Code
7	BoQ	93.7	No	BoQ: A Place is Worth a Bag of Learnable Queries	2024-05-12	Code
8	SegVLAD-FineT (M)	93.1	No	Revisit Anything: Visual Place Recognition via I...	2024-09-26	Code
9	ProGEO	93	Yes	ProGEO: Generating Prompts through Image-Text Co...	2024-06-04	Code
10	SelaVPR	92.8	No	Towards Seamless Adaptation of Pre-trained Model...	2024-02-22	Code
11	EigenPlaces	92.5	No	EigenPlaces: Training Viewpoint Robust Models fo...	2023-08-21	Code
12	BoQ (ResNet-50)	92.4	No	BoQ: A Place is Worth a Bag of Learnable Queries	2024-05-12	Code
13	MixVPR	91.52	No	MixVPR: Feature Mixing for Visual Place Recognit...	2023-03-03	Code
14	CosPlace	90.45	No	Rethinking Visual Geo-localization for Large-Sca...	2022-04-05	Code
15	CosPlace (ResNet-101 2048-D)	90.4	No	Rethinking Visual Geo-localization for Large-Sca...	2022-04-05	Code
16	Patch-NetVLAD	88.7	No	Patch-NetVLAD: Multi-Scale Fusion of Locally-Glo...	2021-03-02	Code
17	AnyLoc-VLAD-DINOv2	87.66	No	AnyLoc: Towards Universal Visual Place Recognition	2023-08-01	Code
18	NetVLAD	86.08	No	NetVLAD: CNN architecture for weakly supervised ...	2015-11-23	Code
19	GCL [trained only on MSLS]	81.94	Yes	Generalized Contrastive Optimization of Siamese ...	2021-03-11	Code
20	DINOv2	78.32	No	DINOv2: Learning Robust Visual Features without ...	2023-04-14	Code
21	DINO	70.13	No	Emerging Properties in Self-Supervised Vision Tr...	2021-04-29	Code
22	CLIP	54.97	No	AnyLoc: Towards Universal Visual Place Recognition	2023-08-01	Code

#1Pair-VPR-pSOTA
95.4
Recall@1· 2024-10-09
Pair-VPR: Place-Aware Pre-training and Contrastive Pair Classification for Visual Place Recognition with Vision Transformers Code
#2Pair-VPR-s
94.7
Recall@1· 2024-10-09
Pair-VPR: Place-Aware Pre-training and Contrastive Pair Classification for Visual Place Recognition with Vision Transformers Code
#3FoL
94.5
Recall@1· 2025-04-14
Focus on Local: Finding Reliable Discriminative Regions for Visual Place Recognition Code
#4QAA-DINOv2-B-8192
94.4
Recall@1· Extra Data· 2025-07-04
Query-Based Adaptive Aggregation for Multi-Dataset Joint Training Toward Universal Visual Place Recognition
#5EffoVPRSOTA
93.9
Recall@1· 2024-05-28
EffoVPR: Effective Foundation Model Utilization for Visual Place Recognition
#6FoL-global
93.9
Recall@1· 2025-04-14
Focus on Local: Finding Reliable Discriminative Regions for Visual Place Recognition Code
#7BoQSOTA
93.7
Recall@1· 2024-05-12
BoQ: A Place is Worth a Bag of Learnable Queries Code
#8SegVLAD-FineT (M)
93.1
Recall@1· 2024-09-26
Revisit Anything: Visual Place Recognition via Image Segment Retrieval Code
#9ProGEO
93
Recall@1· Extra Data· 2024-06-04
ProGEO: Generating Prompts through Image-Text Contrastive Learning for Visual Geo-localization Code
#10SelaVPRSOTA
92.8
Recall@1· 2024-02-22
Towards Seamless Adaptation of Pre-trained Models for Visual Place Recognition Code
#11EigenPlacesSOTA
92.5
Recall@1· 2023-08-21
EigenPlaces: Training Viewpoint Robust Models for Visual Place Recognition Code
#12BoQ (ResNet-50)
92.4
Recall@1· 2024-05-12
BoQ: A Place is Worth a Bag of Learnable Queries Code
#13MixVPRSOTA
91.52
Recall@1· 2023-03-03
MixVPR: Feature Mixing for Visual Place Recognition Code
#14CosPlaceSOTA
90.45
Recall@1· 2022-04-05
Rethinking Visual Geo-localization for Large-Scale Applications Code
#15CosPlace (ResNet-101 2048-D)
90.4
Recall@1· 2022-04-05
Rethinking Visual Geo-localization for Large-Scale Applications Code
#16Patch-NetVLADSOTA
88.7
Recall@1· 2021-03-02
Patch-NetVLAD: Multi-Scale Fusion of Locally-Global Descriptors for Place Recognition Code
#17AnyLoc-VLAD-DINOv2
87.66
Recall@1· 2023-08-01
AnyLoc: Towards Universal Visual Place Recognition Code
#18NetVLADSOTA
86.08
Recall@1· 2015-11-23
NetVLAD: CNN architecture for weakly supervised place recognition Code
#19GCL [trained only on MSLS]
81.94
Recall@1· Extra Data· 2021-03-11
Generalized Contrastive Optimization of Siamese Networks for Place Recognition Code
#20DINOv2
78.32
Recall@1· 2023-04-14
DINOv2: Learning Robust Visual Features without Supervision Code
#21DINO
70.13
Recall@1· 2021-04-29
Emerging Properties in Self-Supervised Vision Transformers Code
#22CLIP
54.97
Recall@1· 2023-08-01
AnyLoc: Towards Universal Visual Place Recognition Code