Cross-View Geo-Localization Based on Dynamic Metric Learning in UAV Scenarios

doi:10.19678/j.issn.1000-3428.0070406

Abstract

Abstract:

Previous studies on cross-view geolocalization have primarily focused on determining whether a query image accurately corresponds to a specific geographic location within a predefined gallery. However, this research paradigm often overlooks the extensive multiscale structural information inherent in the geographic space. To achieve more robust localization, a model must not only capture local architectural details but also understand the spatial relationships among targets reflected through building clusters and environmental features, thereby improving the localization accuracy across different spatial scales. To address these challenges, a multiscale cross-view geo-localization task is proposed and a Multi-Level Campus (ML-Campus) dataset is constructed specifically for this task. The ML-Campus dataset comprises multiview, multisource building images, each annotated with multiscale labels to reflect correlations and continuity across different spatial scales. Based on this dataset, an empirical evaluation of existing cross-view geo-localization methods is conducted, which is used as a benchmark to measure their performance in this context. To further enhance model performance, the proposed Cross-View HAPPIER (CV-HAPPIER) method is employed for training, which strengthens the model's feature representation capabilities across different spatial scales. Extensive experimental results on the ML-Campus dataset demonstrate that the CV-HAPPIER method significantly improves the spatial robustness of cross-view geo-localization retrieval ranking results.

Key words: Unmanned Aerial Vehicle (UAV) scenario, dynamic metric learning, cross-view geo-localization, image retrieval, deep learning

摘要：

现有跨视角地理定位研究主要聚焦于判断查询图像是否准确对应预定义图集中某个特定地理位置, 然而这种研究模式往往忽略了地理空间中固有的大量多空间尺度结构信息。为了实现更为稳健的定位效果, 模型不仅需要捕捉局部建筑细节, 还需理解通过建筑群和环境特征体现的目标之间的空间关系, 从而在不同空间尺度下提高定位准确性。为应对这些挑战, 提出多空间尺度跨视角地理定位任务, 并专门为此任务构建ML-Campus(Multi-Level Campus)数据集。ML-Campus数据集包含多视角、多来源的建筑图像, 并为每个图像标注了多空间尺度标签, 以体现不同空间尺度下的关联性和连续性。基于该数据集, 对现有的跨视角地理定位方法进行了实证评估, 以此为基准衡量其在此背景下的性能表现。为了进一步提升模型性能, 使用提出的CV-HAPPIER(Cross-View HAPPIER)方法进行训练, 以增强模型在不同空间尺度下的特征表示能力。大量在ML-Campus数据集上的实验结果表明, CV-HAPPIER方法显著提升了跨视角地理定位检索排名结果的空间鲁棒性。

关键词: 无人机场景, 动态度量学习, 跨视角地理定位, 图像检索, 深度学习

WANG Xinxin, HU Haifeng, ZHANG Suofei, ZHOU Feifei, GONG Rui. Cross-View Geo-Localization Based on Dynamic Metric Learning in UAV Scenarios[J]. Computer Engineering, 2026, 52(9): 204-217.

王鑫鑫, 胡海峰, 张索非, 周飞飞, 龚锐. 无人机场景下基于动态度量学习的跨视角地理定位[J]. 计算机工程, 2026, 52(9): 204-217.

/ Recommend / Download Citations

URL: https://www.ecice06.com/EN/10.19678/j.issn.1000-3428.0070406

https://www.ecice06.com/EN/Y2026/V52/I9/204

Figures/Tables 13

Fig.1 Comparison between a multi-scale method and a fine-grained method

Fig.2 Spatial partitioning schemes

Fig.3 Architecture for multi-scale cross-view geo-localization method

Fig.4

$ \mathcal{H}$

-AP metric for multi-scale cross-view geo-localization

Fig.5 Cross-view batch processing mechanism

Fig.6 Analysis of hyperparameters λ in Satellite→Drone and Drone→Satellite tasks on ML-Campus dataset

Fig.7 Comparison among single-scale and multi-scale learning on mAP metrics

References 47

1	LIN T Y, CUI Y, BELONGIE S, et al. Learning deep representations for ground-to-aerial geolocalization[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2015: 5007-5015.
2	彭艺, 唐剑, 杨青青. 复杂山区环境下的应急无人机路径规划. 吉林大学学报(理学版), 2025, 63 (2): 585- 594.
	PENG Y , TANG J , YANG Q Q . Emergency UAV path planning in complex mountainous environment. Journal of Jilin University (Science Edition), 2025, 63 (2): 585- 594.
3	WU Y , LIN Y T , DONG X Y , et al. Progressive learning for person re-identification with one example. IEEE Transactions on Image Processing, 2019, 28 (6): 2872- 2881. doi: 10.1109/TIP.2019.2891895
4	YU Q, WANG C, CETINER B, et al. Building information modeling and classification by visual learning at a city scale[EB/OL]. [2024-08-17]. https://arxiv.org/abs/1910.06391.
5	ZHU P F, WEN L, BIAN X, et al. Vision meets drones: a challenge[EB/OL]. [2024-08-17]. https://arxiv.org/abs/1804.07437.
6	ISAAC-MEDINA B K S, POYSER M, ORGANISCIAK D, et al. Unmanned aerial vehicle visual detection and tracking using deep neural networks: a performance benchmark[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops (ICCVW). Washington D.C., USA: IEEE Press, 2021: 1223-1232.
7	GOUEL M, VERMEULEN K, FOURMAUX O, et al. Longitudinal study of an IP geolocation database[EB/OL]. [2024-08-17]. https://arxiv.org/abs/2107.03988.
8	CHEN L P , SUN Y J , SAEED S . Monitoring and predicting land use and land cover changes using remote sensing and GIS techniques—a case study of a hilly area, Jiangle, China. PLoS One, 2018, 13 (7): e0200493. doi: 10.1371/journal.pone.0200493
9	ZHENG L, SHEN L Y, TIAN L, et al. Scalable person re-identification: a benchmark[C]//Proceedings of the IEEE International Conference on Computer Vision (ICCV). Washington D.C., USA: IEEE Press, 2016: 1116-1124.
10	RAMZI E, AUDEBERT N, THOME N, et al. Hierarchical average precision training for pertinent image retrieval[C]//Proceedings of ECCV 2022. Berlin, Germany: Springer, 2022: 250-266.
11	ZHENG Z D, WEI Y C, YANG Y. University-1652: a multi-view multi-source benchmark for drone-based geo-localization[C]//Proceedings of the 28th ACM International Conference on Multimedia. New York, USA: ACM Press, 2020: 1395-1403.
12	CASTALDO F, ZAMIR A, ANGST R, et al. Semantic cross-view matching[C]//Proceedings of the IEEE International Conference on Computer Vision Workshop (ICCVW). Washington D.C., USA: IEEE Press, 2016: 1044-1052.
13	SENLET T, ELGAMMAL A. A framework for global vehicle localization using stereo images and satellite and road maps[C]//Proceedings of the IEEE International Conference on Computer Vision Workshops (ICCV Workshops). Washington D.C., USA: IEEE Press, 2012: 2034-2041.
14	BANSAL M, SAWHNEY H S, CHENG H, et al. Geo-localization of street views with aerial image databases[C]//Proceedings of the 19th ACM International Conference on Multimedia. New York, USA: ACM Press, 2011: 1125-1128.
15	WORKMAN S, JACOBS N. On the location dependence of convolutional neural network features[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Washington D.C., USA: IEEE Press, 2015: 70-78.
16	WORKMAN S, SOUVENIR R, JACOBS N. Wide-area image geolocalization with aerial reference imagery[C]//Proceedings of the IEEE International Conference on Computer Vision (ICCV). Washington D.C., USA: IEEE Press, 2016: 3961-3969.
17	周博文, 李阳, 马鑫骥, 等. 深度学习的跨视角地理定位方法综述. 中国图象图形学报, 2024, 29 (12): 3543- 3563.
	ZHOU B W , LI Y , MA X J , et al. A review of deep learning-based cross-view geolocalization methods. Journal of Image and Graphics, 2024, 29 (12): 3543- 3563.
18	DEUSER F, HABEL K, OSWALD N. Sample4Geo: hard negative sampling for cross-view geo-localisation[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Washington D.C., USA: IEEE Press, 2024: 16801-16810.
19	HADSELL R, CHOPRA S, LECUN Y. Dimensionality reduction by learning an invariant mapping[C]//Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Washington D.C., USA: IEEE Press, 2006: 1735-1742.
20	DENG W J, ZHENG L, YE Q X, et al. Image-image domain adaptation with preserved self-similarity and domain-dissimilarity for person re-identification[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D.C., USA: IEEE Press, 2018: 994-1003.
21	HU S X, FENG M D, NGUYEN R M H, et al. CVM-Net: cross-view matching network for image-based ground-to-aerial geo-localization[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D.C., USA: IEEE Press, 2018: 7258-7267.
22	ARANDJELOVIC R, GRONAT P, TORII A, et al. NetVLAD: CNN architecture for weakly supervised place recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2016: 5297-5307.
23	LIU L, LI H D. Lending orientation to neural networks for cross-view geo-localization[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2020: 5617-5626.
24	SHI Y J, YU X, LIU L, et al. Optimal feature transport for cross-view image geo-localization[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Palo Alto, USA: AAAI Press, 2020: 11990-11997.
25	SHI Y J, LIU L, YU X, et al. Spatial-aware feature aggregation for image based cross-view geo-localization[C]//Proceedings of the 33rd International Conference on Neural Information Processing Systems. New York, USA: ACM Press, 2019: 10090-10100.
26	SHI Y J, YU X, CAMPBELL D, et al. Where am I looking at? Joint location and orientation estimation by cross-view matching[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2020: 4063-4071.
27	VO N N, HAYS J. Localizing and orienting street views using overhead imagery[C]//Proceedings of European Conference on Computer Vision. Berlin, Germany: Springer, 2016: 494-509.
28	SCHROFF F, KALENICHENKO D, PHILBIN J. FaceNet: a unified embedding for face recognition and clustering[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2015: 815-823.
29	LI P K , PAN P B , LIU P , et al. Hierarchical temporal modeling with mutual distance matching for video based person re-identification. IEEE Transactions on Circuits and Systems for Video Technology, 2021, 31 (2): 503- 511. doi: 10.1109/TCSVT.2020.2988034
30	ZHENG Z D, ZHENG L, YANG Y. Unlabeled samples generated by GAN improve the person re-identification baseline in vitro[C]//Proceedings of the IEEE International Conference on Computer Vision (ICCV). Washington D.C., USA: IEEE Press, 2017: 3774-3782.
31	ZHENG Z D , ZHENG L , GARRETT M , et al. Dual-path convolutional image-text embeddings with instance loss. ACM Transactions on Multimedia Computing, Communications, and Applications, 2020, 16 (2): 1- 23.
32	WANG T Y , ZHENG Z D , YAN C G , et al. Each part matters: local patterns facilitate cross-view geo-localization. IEEE Transactions on Circuits and Systems for Video Technology, 2022, 32 (2): 867- 879. doi: 10.1109/TCSVT.2021.3061265
33	TRAMÈR F, CARLINI N, BRENDEL W, et al. On adaptive attacks to adversarial example defenses[EB/OL]. [2024-08-17]. https://arxiv.org/abs/2002.08347.
34	LIU Z W, LUO P, QIU S, et al. DeepFashion: powering robust clothes recognition and retrieval with rich annotations[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2016: 1096-1104.
35	SUN Y F, ZHU Y K, ZHANG Y H, et al. Dynamic metric learning: towards a scalable metric space to accommodate multiple semantic scales[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2021: 5389-5398.
36	ROLINEK M, MUSIL V, PAULUS A, et al. Optimizing rank-based metrics with blackbox differentiation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2020: 7617-7627.
37	BROWN A, XIE W D, KALOGEITON V, et al. Smooth$ \mathcal{H}$-AP: smoothing the path towards large-scale image retrieval[C]// Proceedings of ECCV 2020. Berlin, Germany: Springer, 2020: 677-694.
38	RAMZI E, THOME N, RAMBOUR C, et al. Robust and decomposable average precision for image retrieval[EB/OL]. [2024-08-17]. https://arxiv.org/abs/2110.01445.
39	赵继达, 甄国涌, 储成群. 基于YOLOv8的无人机图像目标检测算法. 计算机工程, 2024, 50 (4): 113- 120.
	ZHAO J D , ZHEN G Y , CHU C Q . Unmanned aerial vehicle image target detection algorithm based on YOLOv8. Computer Engineering, 2024, 50 (4): 113- 120.
40	王林, 赵莉, 王无为. 高动态场景下无人机空对空目标检测. 计算机工程, 2024, 50 (12): 265- 275. doi: 10.19678/j.issn.1000-3428.0068620
	WANG L , ZHAO L , WANG W W . Air-to-air target detection of unmanned aerial vehicles under high dynamic scenarios. Computer Engineering, 2024, 50 (12): 265- 275. doi: 10.19678/j.issn.1000-3428.0068620
41	王青旺, 欧阳俊林, 靳鹏程, 等. 基于跨模态特征解耦与重聚焦的无人机道路场景语义分割. 遥感技术与应用, 2025, 40 (4): 864- 874.
	WANG Q W , OUYANG J L , JIN P C , et al. Semantic segmentation of UAV road scenes based on cross-modal feature decoupling and refocusing. Remote Sensing Technology and Application, 2025, 40 (4): 864- 874.
42	JÄRVELIN K , KEKÄLÄINEN J . Cumulated gain-based evaluation of IR techniques. ACM Transactions on Information Systems, 2002, 20 (4): 422- 446. doi: 10.1145/582415.582418
43	ZHOU B L , LAPEDRIZA A , KHOSLA A , et al. Places: a 10 million image database for scene recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40 (6): 1452- 1464. doi: 10.1109/TPAMI.2017.2723009
44	HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2016: 770-778.
45	DENG J, DONG W, SOCHER R, et al. ImageNet: a large-scale hierarchical image database[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Washington D.C., USA: IEEE Press, 2009: 248-255.
46	ZHAI A, WU H. Making classification competitive for deep metric learning[EB/OL]. [2024-08-17]. https://arxiv.org/abs/1811.12649v1.
47	WANG X, HAN X, HUANG W, et al. Multi-similarity loss with general pair weighting for deep metric learning[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2019: 5017-5025.

[1]	FAN Shuai, CHEN Yonghui, QIAO Xu, XIAO Liwei. 2D Obstacle Flow Field Prediction Based on Geometry Adaptation and Multi-Scale Gated Fusion [J]. Computer Engineering, 2026, 52(8): 376-389.
[2]	LI Zheng, LI Zhixiao, QIN Jinlei, GUO Changzhen. Multiple Loads Forecasting Based on Pattern Cross-Correlation and Temporal Patch Association Mechanism [J]. Computer Engineering, 2026, 52(8): 315-326.
[3]	ZHAO Yingying, ZHU Shuaishuai. Research Review of Knowledge Graph Construction Technology [J]. Computer Engineering, 2026, 52(8): 84-100.
[4]	WANG Zhongmei, NIE Pengxuan, LIU Jianhua, WU Haibo, ZHENG Liang. Bearing Fault Diagnosis Based on Multiple-Constraint Modal-Invariant Graph Convolutional Fusion Network [J]. Computer Engineering, 2026, 52(8): 422-430.
[5]	HU Caifu, WEI Bo, REN Ruibin. Network Traffic Classification Based on Incremental Learning from the Perspective of Space Optimization [J]. Computer Engineering, 2026, 52(8): 138-148.
[6]	WEI Fangda, LIU Miao, SUN Yi, WANG Jing, ZHAO Shenghui. Review of Audio and Video Deepfakes and Counterfeit Detection [J]. Computer Engineering, 2026, 52(7): 40-55.
[7]	LIANG Ziyi, WANG Zihao, LIU Tianquan, LI Liping, ZHU Yuanfei, LU Cunyue. PASRGAN: A Super-Resolution Pathological Image Generation Algorithm Based on Parallel Attention Mechanism [J]. Computer Engineering, 2026, 52(7): 189-198.
[8]	FENG Jiaqi, GAO Jian. Review of Information Extraction in the Field of Cyber Threat Intelligence [J]. Computer Engineering, 2026, 52(7): 56-75.
[9]	HU Kangyuan, GUO Tao, MU Nan. Image Inpainting Method for Cultural Relics Based on Self-Attention Mechanism and Dynamic Masking Mechanism [J]. Computer Engineering, 2026, 52(6): 179-188.
[10]	XIAO Zeqiu, LI Yong, WANG Xia. Prediction of Blood Glucose Concentration in Diabetic Patients Based on PBI-CLA Model [J]. Computer Engineering, 2026, 52(6): 382-390.
[11]	LI Yiran, NIE Hongbin, YANG Ziqian, BIAN Chunjiang. Detection Method for Space Dim and Weak Group Targets in Dense Star Fields [J]. Computer Engineering, 2026, 52(6): 149-159.
[12]	SUN Haifeng, YAO Junping, LI Xiaojun, LIU Yanfei, GU Hongyang. Review of Deep Learning Methods for Short-Term Action Anticipation [J]. Computer Engineering, 2026, 52(6): 31-52.
[13]	LI Xuexiang, ZHENG Yongli, ZHANG Yize, DUAN Pengsong. Review of Traffic Analysis Methods Based on Machine Learning and Pre-trained Model [J]. Computer Engineering, 2026, 52(6): 53-67.
[14]	XU Minchen, QU Dan, SI Nianwen, PENG Sisi, CHEN Yaqi. Technologies for Detecting Disinformation in Social Media: A Comprehensive Review [J]. Computer Engineering, 2026, 52(5): 60-80.
[15]	LI Hui, LIU Jiayu, XU Yaping. Review on Deep Learning Model Architectures and Performance Evaluation Methods for Medical Image Segmentation [J]. Computer Engineering, 2026, 52(5): 81-94.

Please choose a citation manager

Content to export