3D-LENS: A 3D Lifting-based Elevated Novel-view Synthesis method for Single-View Aerial-Ground Re-Identification

2026-04-29

AI Summary

This work addresses the challenge of feature occlusion and distortion in aerial-to-ground person re-identification under single-view training, caused by significant viewpoint domain discrepancies. It introduces, for the first time, a single-view cross-view re-identification task that does not require paired data from the target domain. The proposed 3D-LENS framework leverages large-scale 3D mesh reconstruction to synthesize geometrically consistent novel viewpoints and integrates adversarial representation learning to enhance generalization. Notably, the method operates without category-specific templates, preserves fine-grained details, and ensures multi-view consistency. Evaluated in real-world scenarios such as wilderness search-and-rescue, it substantially outperforms existing 2D generation-based and template-driven 3D approaches, achieving state-of-the-art performance.

Abstract

Aerial-Ground Re-Identification (AG-ReID) is constrained by the viewpoint-domain gap, as drastic viewpoint disparities occlude or distort discriminative features, making cross-viewpoint image retrieval challenging. While existing methods rely on paired cross-view annotations, real-world deployments, such as wilderness search-and-rescue (SAR), often lack target-domain data, requiring retrieval from ground-level references alone. To our knowledge, we are the first to address this challenge by formalizing the Single-View AG-ReID (SV AG-ReID) setting, where models trained on a single real viewpoint must generalize to an unseen viewpoint. We propose 3D Lifting-based Elevated Novel-view Synthesis (3D-LENS), a unified framework combining geometrically-consistent novel view synthesis that leverages large-scale 3D mesh reconstruction, with a robust representation learning scheme to mitigate synthetic-to-real bias. Unlike 2D generative baselines that suffer from geometric inconsistencies or prior 3D methods that are restricted to class-specific templates, our approach ensures view-consistent synthesis across diverse categories without predefined templates that fail to capture fine-grained details, such as carried objects. Extensive experiments demonstrate that our method achieves state-of-the-art performance on SV AG-ReID scenarios. Code and data will be released at https://github.com/TurtleSmoke/3D-LENS.

Problem

• Aerial-Ground Re-Identification
• Single-View Re-Identification
• Novel-view Synthesis
• Cross-viewpoint Retrieval
• Viewpoint-domain Gap

Innovation

• 3D novel-view synthesis
• single-view re-identification
• geometric consistency
• aerial-ground ReID
• synthetic-to-real generalization