Skip to content

Latest commit

 

History

History
 
 

State-of-the-art

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 

Awesome Geo-localization

University-1652 Dataset

Drone <-> Satellite

Methods R@1 AP R@1 AP Reference
Drone -> Satellite Satellite -> Drone
Contrastive Loss 52.39 57.44 63.91 52.24
Triplet Loss (margin=0.3) 55.18 59.97 63.62 53.85
Triplet Loss (margin=0.5) 53.58 58.60 64.48 53.15
Weighted Soft Margin Triplet Loss 53.21 58.03 65.62 54.47 Liu L, Li H. Lending orientation to neural networks for cross-view geo-localization[C]. CVPR, 2019: 5624-5633. [Paper]
Instance Loss 58.23 62.91 74.47 59.45 Zheng Z, Zheng L, Garrett M, et al. Dual-Path Convolutional Image-Text Embedding with Instance Loss. TOMM 2020. [Paper]
Instance Loss + Verification Loss 61.30 65.68 75.04 62.87 Zheng Z, Zheng L, Yang Y. A discriminatively learned cnn embedding for person reidentification[J]. TOMM, 2017, 14(1): 1-20. [Paper] [Code]
Instance Loss + GeM Pooling 65.32 69.61 79.03 65.35 Radenović, Filip, Giorgos Tolias, and Ondřej Chum. "Fine-tuning CNN image retrieval with no human annotation." TPAMI (2018): 1655-1668.
Instance Loss + Weighted Soft Margin Triplet Loss 65.93 70.18 76.03 66.36
RK-Net (USAM) 66.13 70.23 80.17 65.76 Lin J, Zheng Z, Zhong Z, Luo Z, Li S, Yang Y, Sebe N. Joint Representation Learning and Keypoint Detection for Cross-view Geo-localization. TIP 2022. [Paper] [Code]
LCM (ResNet-50) 66.65 70.82 79.89 65.38 Ding L, Zhou J, Meng L, et al. A Practical Cross-View Image Matching Method between UAV and Satellite for UAV-Based Geo-Localization[J]. Remote Sensing, 2021, 13(1): 47. [Paper]
Instance Loss + GNN ReRanking 70.30 74.11 - - Zhang, Xuanmeng, Minyue Jiang, Zhedong Zheng, Xiao Tan, Errui Ding, and Yi Yang. "Understanding Image Retrieval Re-Ranking: A Graph Neural Network Perspective." arXiv 2020. [Paper][Code]
Instance Loss + USAM + SAFA 72.19 75.79 83.23 71.77
LPN 75.93 79.14 86.45 74.79 Tingyu W, Zhedong Z, Chenggang Y, and Yi Y. Each Part Matters: Local Patterns Facilitate Cross-view Geo-localization. TCSVT 2021. [Paper] [Code]
Instance Loss + Weighted Soft Margin Triplet Loss + LPN 76.29 79.46 81.74 73.58
Instance Loss + Verification Loss + LPN 77.08 80.18 85.02 73.80
Instance Loss + USAM + LPN 77.60 80.55 86.59 75.96
PCL 79.47 83.63 87.69 78.51 Xiaoyang Tian, Jie Shao, Deqiang Ouyang, and Heng Tao Shen. UAV-Satellite View Synthesis for Cross-view Geo-Localization. TCSVT 2021. [Paper]
FSRA (k=1) 82.25 84.82 87.87 81.53 Ming Dai, Jianhong Hu, Jiedong Zhuang, Enhui Zheng. A Transformer-Based Feature Segmentation and Region Alignment Method For UAV-View Geo-Localization. TCSVT 2022. [Paper] [Code]
FSRA (k=3) 84.51 86.71 88.45 83.37

Ground <-> Satellite

Methods Training Set R@1 AP R@1 AP Reference
Ground -> Satellite Satellite -> Ground
Instance Loss Satellite + Ground 0.62 1.60 0.86 1.00 Zheng Z, Zheng L, Garrett M, et al. Dual-Path Convolutional Image-Text Embedding with Instance Loss. TOMM 2020. [Paper]
Instance Loss Satellite + Drone + Ground 1.28 2.29 1.57 1.52
Instance Loss Satellite + Drone + Ground + Google Image 1.20 2.52 1.14 1.41
LPN Satellite + Ground 0.74 1.83 1.43 1.31 Tingyu W, Zhedong Z, Chenggang Y, and Yi Y. Each Part Matters: Local Patterns Facilitate Cross-view Geo-localization. TCSVT 2021. [Paper] [Code]
LPN Satellite + Drone + Ground 0.81 2.21 1.85 1.66 Tingyu W, Zhedong Z, Chenggang Y, and Yi Y. Each Part Matters: Local Patterns Facilitate Cross-view Geo-localization. TCSVT 2021. [Paper] [Code]
PCLD Satellite + Drone + Ground 9.15 14.16 - - Zeng, Z., Wang, Z., Yang, F., & Satoh, S. I. (2022). Geo-Localization via Ground-to-Satellite Cross-View Image Retrieval. IEEE Transactions on Multimedia. [Paper]

cvusa Dataset

Methods R@1 R@5 R@10 R@Top1 Reference
Workman - - - 34.40 Scott Workman, Richard Souvenir, and Nathan Jacobs. ICCV 2015. Wide-area image geolocalization with aerial reference imagery [Paper]
Zhai - - - 43.20 Menghua Zhai, Zachary Bessinger, Scott Workman, and Nathan Jacobs. CVPR 2017. Predicting ground-level scene layout from aerial imagery.[Paper]
Vo - - - 63.70 Nam N Vo and James Hays. ECCV 2016. Localizing and orienting street views using overhead imagery
CVM-Net 18.80 44.42 57.47 91.54 Sixing Hu, Mengdan Feng, Rang MH Nguyen, and Gim Hee Lee. CVPR 2018. CVM-net:Cross-view matching network for image-based ground-to-aerial geo-localization. [Paper]
Orientation* 27.15 54.66 67.54 93.91 Liu Liu and Hongdong Li. CVPR 2019. Lending Orientation to Neural Networks for Cross-view Geo-localization [Paper]
Siam-FCANet - - - 98.3 Sudong C, Yulan G, Salman K, et al. Ground-to-Aerial Image Geo-Localization With a Hard Exemplar Reweighting Triplet Loss. ICCV 2019. [Paper]
Feature Fusion 48.75 - 81.27 95.98 Krishna Regmi, Mubarak Shah, et al. Bridging the Domain Gap for Ground-to-Aerial Image Matching. ICCV 2019. [Paper]
Instance Loss 43.91 66.38 74.58 91.78 Zheng Z, Zheng L, Garrett M, et al. Dual-Path Convolutional Image-Text Embedding with Instance Loss. TOMM 2020. [Paper] [Code]
RK-Net (USAM) 52.50 - - 96.52 Lin J, Zheng Z, Zhong Z, Luo Z, Li S, Yang Y, Sebe N. Joint Representation Learning and Keypoint Detection for Cross-view Geo-localization. TIP 2022. [Paper] [Code]
CVFT 61.43 84.69 90.49 99.02 Shi Y, Yu X, Liu L, et al. Optimal Feature Transport for Cross-View Image Geo-Localization. AAAI 2020. [Paper]
MS Attention w DataAug 75.95 91.90 95.00 99.42 Rodrigues, Royston, and Masahiro Tani. "Are These From the Same Place? Seeing the Unseen in Cross-View Image Geo-Localization." WACV 2021. [Paper]
LPN 85.79 95.38 96.98 99.41 Tingyu Wang, Zhedong Zheng, Chenggang Yan, and Yi, Yang. Each Part Matters: Local Patterns Facilitate Cross-view Geo-localization. TCSVT 2021. [Paper] [Code]
SAFA 89.84 96.93 98.14 99.64 Yujiao Shi, Liu Liu, Xin Yu, et al. Spatial-Aware Feature Aggregation for Cross-View Image based Geo-Localization. NIPS 2019. [Paper]
SAFA + USAM 90.16 - - 99.67 Lin J, Zheng Z, Zhong Z, Luo Z, Li S, Yang Y, Sebe N. Joint Representation Learning and Keypoint Detection for Cross-view Geo-localization. TIP 2022. [Paper] [Code]
LPN + USAM 91.22 - - 99.67 Lin J, Zheng Z, Zhong Z, Luo Z, Li S, Yang Y, Sebe N. Joint Representation Learning and Keypoint Detection for Cross-view Geo-localization. TIP 2022. [Paper] [Code]
DSM 91.96 97.50 98.54 99.67 Yujiao Shi, Xin Yu, Dylan Campbell, and Hongdong Li. "Where am i looking at? joint location and orientation estimation by cross-view matching." CVPR 2020. [Paper] [Code]
Toker etal. 92.56 97.55 98.33 99.67 Aysim Toker, Qunjie Zhou, Maxim Maximov, Laura Leal-Taixé. Coming Down to Earth: Satellite-to-Street View Synthesis for Geo-Localization. CVPR 2021 [Paper]
SAFA + LPN 92.83 98.00 98.85 99.78 Tingyu Wang, Zhedong Zheng, Chenggang Yan, and Yi, Yang. Each Part Matters: Local Patterns Facilitate Cross-view Geo-localization. TCSVT 2021. [Paper] [Code]
Polar-EgoTR 94.05 98.27 98.99 99.67 Hongji Yang, Xiufan Lu, Yingying Zhu. Cross-view Geo-localization with Layer-to-Layer Transformer. Nips 2021 [Paper] [Code]
TransGeo 94.08 98.36 99.04 99.77 Sijie Zhu, Mubarak Shah, Chen Chen. TransGeo: Transformer Is All You Need for Cross-view Image Geo-localization. CVPR 2022 [Paper] [Code]
*: The method utilizes extra orientation information as input.

cvact Dataset

Methods R@1 R@5 R@10 R@Top1 Reference
CVM-Net 20.15 45.00 56.87 87.57 Sixing Hu, Mengdan Feng, Rang MH Nguyen, and Gim Hee Lee. CVPR 2018. CVM-net:Cross-view matching network for image-based ground-to-aerial geo-localization. [Paper]
Instance Loss 31.20 53.64 63.00 85.27 Zheng Z, Zheng L, Garrett M, et al. Dual-Path Convolutional Image-Text Embedding with Instance Loss. TOMM 2020. [Paper] [Code]
RK-Net (USAM) 40.53 - - 89.12 Lin J, Zheng Z, Zhong Z, Luo Z, Li S, Yang Y, Sebe N. Joint Representation Learning and Keypoint Detection for Cross-view Geo-localization. TIP 2022. [Paper] [Code]
Orientation* 46.96 68.28 75.48 92.04 Liu Liu and Hongdong Li. CVPR 2019. Lending Orientation to Neural Networks for Cross-view Geo-localization [Paper]
CVFT 61.05 81.33 86.52 95.93 Shi Y, Yu X, Liu L, et al. Optimal Feature Transport for Cross-View Image Geo-Localization. AAAI 2020. [Paper]
MS Attention w DataAug 73.19 90.39 93.38 97.45 Rodrigues, Royston, and Masahiro Tani. "Are These From the Same Place? Seeing the Unseen in Cross-View Image Geo-Localization." WACV 2021. [Paper]
LPN 79.99 90.63 92.56 97.03 Tingyu Wang, Zhedong Zheng, Chenggang Yan, and Yi, Yang. Each Part Matters: Local Patterns Facilitate Cross-view Geo-localization. TCSVT 2021. [Paper] [Code]
SAFA 81.03 92.80 94.84 98.17 Yujiao Shi, Liu Liu, Xin Yu, et al. Spatial-Aware Feature Aggregation for Cross-View Image based Geo-Localization. NIPS 2019. [Paper]
LPN + USAM 82.02 - - 98.18 Lin J, Zheng Z, Zhong Z, Luo Z, Li S, Yang Y, Sebe N. Joint Representation Learning and Keypoint Detection for Cross-view Geo-localization. TIP 2022. [Paper] [Code]
SAFA + USAM 82.40 - - 98.00 Lin J, Zheng Z, Zhong Z, Luo Z, Li S, Yang Y, Sebe N. Joint Representation Learning and Keypoint Detection for Cross-view Geo-localization. TIP 2022. [Paper] [Code]
DSM 82.49 92.44 93.99 97.32 Yujiao Shi, Xin Yu, Dylan Campbell, and Hongdong Li. "Where am i looking at? joint location and orientation estimation by cross-view matching." CVPR 2020. [Paper] [Code]
Toker etal. 83.28 93.57 95.42 98.22 Aysim Toker, Qunjie Zhou, Maxim Maximov, Laura Leal-Taixé. Coming Down to Earth: Satellite-to-Street View Synthesis for Geo-Localization. CVPR 2021 [Paper]
SAFA + LPN 83.66 94.14 95.92 98.41 Tingyu Wang, Zhedong Zheng, Chenggang Yan, and Yi, Yang. Each Part Matters: Local Patterns Facilitate Cross-view Geo-localization. TCSVT 2021. [Paper] [Code]
Polar-EgoTR 84.89 94.59 95.96 98.37 Hongji Yang, Xiufan Lu, Yingying Zhu. Cross-view Geo-localization with Layer-to-Layer Transformer. Nips 2021 [Paper] [Code]
TransGeo 84.95 94.14 95.78 98.37 Sijie Zhu, Mubarak Shah, Chen Chen. TransGeo: Transformer Is All You Need for Cross-view Image Geo-localization. CVPR 2022 [Paper] [Code]
*: The method utilizes extra orientation information as input.