Recurrent Cross-View Object Geo-Localization
Authors:
Preprint
Abstract
Cross-view object geo-localization (CVOGL) aims to determine the location of a specific object in high-resolution satellite imagery given a query image with a point prompt. Existing approaches treat CVOGL as a one-shot detection task, directly regressing object locations from cross-view information aggregation, but they are vulnerable to feature noise and lack mechanisms for error correction. In this paper, we propose ReCOT, a Recurrent Cross-view Object geo-localization Transformer, which reformulates CVOGL as a recurrent localization task. ReCOT introduces a set of learnable tokens that encode task-specific intent from the query image and prompt embeddings, and iteratively attend to the reference features to refine the predicted location. To enhance this recurrent process, we incorporate two complementary modules: (1) a SAM-based knowledge distillation strategy that transfers segmentation priors from the Segment Anything Model (SAM) to provide clearer semantic guidance without additional inference cost, and (2) a Reference Feature Enhancement Module (RFEM) that introduces a hierarchical attention to emphasize object-relevant regions in the reference features. Extensive experiments on standard CVOGL benchmarks demonstrate that ReCOT achieves state-of-the-art (SOTA) performance while reducing parameters by 60% compared to previous SOTA approaches.Keywords
All Related Versions
This publication has 16 references indexed in Scilit:
- Object-Level Cross-View Geolocalization With Location Enhancement and Multihead Cross AttentionIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2025
- VAGeo: View-specific Attention for Cross-View Object Geo-Localization2025
- Semantic enrichment of BIM with IndoorGML for quadruped robot navigation and automated 3D scanningAutomation in Construction, 2024
- MRF3Net: An Infrared Small Target Detection Network Using Multireceptive Field Perception and Effective Feature FusionIEEE Transactions on Geoscience and Remote Sensing, 2024
- Multistage Enhancement Network for Tiny Object Detection in Remote Sensing ImagesIEEE Transactions on Geoscience and Remote Sensing, 2024
- Cross-View Object Geo-Localization in a Local Region With Satellite ImageryIEEE Transactions on Geoscience and Remote Sensing, 2023
- Trinity-Net: Gradient-Guided Swin Transformer-Based Remote Sensing Image Dehazing and BeyondIEEE Transactions on Geoscience and Remote Sensing, 2023
- Geographic Semantic Network for Cross-View Image Geo-LocalizationIEEE Transactions on Geoscience and Remote Sensing, 2021
- Iterative Residual Refinement for Joint Optical Flow and Occlusion Estimation2019
- Exploiting SAR and VHR Optical Images to Quantify Damage Caused by the 2003 Bam EarthquakeIEEE Transactions on Geoscience and Remote Sensing, 2008