Reliability-Aware 3D Geometric Injection for Universal Person Re-identification

2026-07-21

AI Summary

This work addresses the challenges of generalizable person re-identification under complex scenarios such as occlusion, clothing change, and cross-modality, where 2D representations suffer from high ambiguity and lack depth and topological awareness. Direct incorporation of monocular 3D priors is often hindered by geometric noise. To overcome this, we propose UniGeo, a novel framework that decouples monocular 3D human pose estimation into two stages: structural extraction and dynamic utilization. A consistency-aware reliability gating mechanism filters out noisy estimates, and the denoised joint topology is adaptively fused in a residual manner. This approach maintains competitive performance on standard benchmarks while significantly enhancing robustness and generalizability in challenging real-world conditions.

Abstract

Universal person re-identification (ReID) aims to retrieve pedestrian identities across diverse real-world scenarios, including severe occlusions, clothing changes, and cross-modality shifts, within a unified model. However, existing 2D representations fundamentally struggle with spatial ambiguities due to a lack of depth and topological awareness, while naively introducing monocular 3D priors often causes severe negative transfer due to geometric estimation noise under extreme visual degradation. To safely harness the clothing-invariant and canonical structural properties of 3D geometry, we propose UniGeo, a Universal Monocular 3D-Enhanced ReID framework driven by a Consistency-Aware Reliability Gate and Dual-Stream Residual Fusion. Specifically, the processing of 3D information is strategically decoupled into geometric extraction and dynamic utilization. To provide pure structural compensation, we project monocular 3D parameters into kinematic joint representations, explicitly capturing instance-level geometric topology to resolve appearance-based ambiguities. To robustly incorporate these cues without perturbing the reliable 2D feature space, we isolate the 3D prior as a late-stage structural residual; modulated by the consistency-aware gate, this mechanism adaptively filters geometric noise and enables controlled fallback to the pure 2D baseline. Extensive experiments show that our method improves challenging, structure-sensitive scenarios while preserving competitive performance on clean domains. Code is available at https://github.com/BohanSu/UniGeo.

Problem

• person re-identification
• 3D geometry
• spatial ambiguity
• negative transfer
• visual degradation

Innovation

• 3D geometric injection
• reliability-aware gating
• universal person re-identification
• monocular 3D prior
• residual fusion