Unsupervised Domain Adaptation for Sim-to-Real Object Pose Estimation with Contrastive Alignment and Pseudo-Label Refinement

2026-06-19

AI Summary

This work addresses the limitations of existing unsupervised domain adaptation methods for object pose estimation, which often overlook pose-sensitive features and struggle to preserve geometric structure. To overcome these challenges, the authors propose a local region-focused contrastive alignment mechanism, integrated with intermediate feature-guided cross-domain image pairing and a pseudo-label refinement strategy under consistency constraints. This approach effectively aligns geometry-aware features between synthetic and real domains, thereby enhancing the stability and accuracy of pose predictions in the target domain. Evaluated on multiple mainstream pose estimation benchmarks, the proposed method significantly outperforms current unsupervised domain adaptation approaches, achieving state-of-the-art performance.

Abstract

Unsupervised domain adaptation (UDA) enables robust transfer of knowledge from simulated to real environments while exploiting a subset of unlabeled target data to improve real-world performance. Existing UDA methods for Object pose estimation often rely on global feature matching, multi-stage larger frameworks, or image translation pipelines, which tend to overlook the pose-specific information embedded in feature representations. To bridge this limitation, we introduce CAPLR that targets the adaptation of pose-sensitive features in localized regions, ensuring that domain alignment preserves the geometric cues essential for accurate pose estimation. CAPLR achieves UDA with three key components: (1) Efficient Cross-Domain Pairing strategy leveraging intermediate features to identify pose similar image pairs across domains without supervision; (2) Contrastive Alignment to perform feature alignment at localised regions in both intermediate and task-specific representations; and (3) Consistency-Based Pseudo-Label Refinement to improve reliability by encouraging stable target predictions. Extensive experiments demonstrate that CAPLR achieves state-of-the-art performance across multiple well-known object pose estimation benchmarks featuring diverse and challenging scenarios.

Problem

• Unsupervised Domain Adaptation
• Object Pose Estimation
• Sim-to-Real
• Pose-Specific Features
• Domain Alignment

Innovation

• Unsupervised Domain Adaptation
• Object Pose Estimation
• Contrastive Alignment
• Pseudo-Label Refinement
• Sim-to-Real Transfer