DiffusionUavLoc: Visually Prompted Diffusion for Cross-View UAV Localization

2025-11-09

AI Summary

In GNSS-denied environments, UAV localization suffers from significant geometric and appearance domain shifts between satellite imagery and oblique aerial views; existing methods rely on complex architectures, textual prompts, or extensive annotations, limiting generalizability. This paper proposes a vision-prompted cross-view localization framework: we introduce the first image-conditioned, text-free diffusion model, integrated with a VAE for unified representation learning and a training-free geometric renderer to synthesize pseudo-satellite images. Cross-view matching is achieved efficiently via fixed-timestep feature extraction and cosine similarity. Our method requires no textual prompts, minimal annotations, and avoids intricate network designs. Evaluated on University-1652 and SUES-200, it achieves state-of-the-art performance—particularly robust in satellite-to-UAV retrieval—demonstrating strong effectiveness and generalization.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Intelligent Robots: Localization, Mapping, and Navigation
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

With the rapid growth of the low-altitude economy, unmanned aerial vehicles (UAVs) have become key platforms for measurement and tracking in intelligent patrol systems. However, in GNSS-denied environments, localization schemes that rely solely on satellite signals are prone to failure. Cross-view image retrieval-based localization is a promising alternative, yet substantial geometric and appearance domain gaps exist between oblique UAV views and nadir satellite orthophotos. Moreover, conventional approaches often depend on complex network architectures, text prompts, or large amounts of annotation, which hinders generalization. To address these issues, we propose DiffusionUavLoc, a cross-view localization framework that is image-prompted, text-free, diffusion-centric, and employs a VAE for unified representation. We first use training-free geometric rendering to synthesize pseudo-satellite images from UAV imagery as structural prompts. We then design a text-free conditional diffusion model that fuses multimodal structural cues to learn features robust to viewpoint changes. At inference, descriptors are computed at a fixed time step t and compared using cosine similarity. On University-1652 and SUES-200, the method performs competitively for cross-view localization, especially for satellite-to-drone in University-1652.Our data and code will be published at the following URL: https://github.com/liutao23/DiffusionUavLoc.git.

Problem

• Addresses UAV localization failure in GNSS-denied environments
• Bridges geometric and appearance gaps between UAV and satellite views
• Reduces dependency on text prompts and complex annotation requirements

Innovation

• Training-free geometric rendering for pseudo-satellite images
• Text-free conditional diffusion model for feature learning
• VAE-based unified representation with cosine similarity comparison