A backbone network mixed with PVT and mixer for cross-view image geo-localization

Anning Ni; Yan Gao

doi:10.1117/12.3032814

5 July 2024 A backbone network mixed with PVT and mixer for cross-view image geo-localization

Anning Ni, Yan Gao

Proceedings Volume 13184, Third International Conference on Electronic Information Engineering and Data Processing (EIEDP 2024); 131841O (2024) https://doi.org/10.1117/12.3032814
Event: 3rd International Conference on Electronic Information Engineering and Data Processing (EIEDP 2024), 2024, Kuala Lumpur, Malaysia

Abstract

In the task of cross-view image-based geolocation, many existing methods rely on coordinate transformations, which fail to expand the perception across dimensions and establish global correlations effectively. Moreover, these methods often require substantial computational resources, and the parameterization of models lacks effective control. This paper proposes a novel approach to cross-view geo-visual localization. Compared to existing methods, it achieves a retrieval accuracy of R@1 reaching 95.42% on the CVUSA dataset while reducing the parameter count by approximately 31.62% compared to TransGeo. Importantly, it does not require the assumption that the center positions of any query images are perfectly aligned. To expedite model inference and reduce parameter requirements, the algorithm's feature extraction backbone employs the PVT structure. Furthermore, it utilizes multi-scale feature maps for subsequent inference, which balances global and local features spatially. However, CNN methods tend to overlook long-range relationships between pixels or patches. Hence, the Mixer method is introduced, which utilizes spatial channel mixing algorithms to obtain more robust global image vector descriptors. Additionally, the unique flexibility of this module ensures long-term contextual relationships without the need for assumptions regarding feature alignment coordinate transformations.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Anning Ni and Yan Gao "A backbone network mixed with PVT and mixer for cross-view image geo-localization", Proc. SPIE 13184, Third International Conference on Electronic Information Engineering and Data Processing (EIEDP 2024), 131841O (5 July 2024); https://doi.org/10.1117/12.3032814

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available