TriCons-Pose: Triangle-Invariant Geometric Consistency Learning for Category-Level Object Pose Estimation

2026-07-12

AI Summary

Current category-level object pose estimation methods suffer from geometric instability in correspondence learning under shape variation and occlusion, leading to fragile performance. To address this issue, this work proposes a Structurally Consistent Keypoint Detector (SCKD) and a Pose-Invariant Geometric Aggregator (PIGA), which jointly learn pose-invariant geometric descriptors through triangle-based geometric consistency. The approach integrates normalized pairwise distance matching, triangle geometric constraints, and a local-global attention mechanism to enhance the geometric stability of correspondences. As a result, the method achieves significantly more accurate and robust category-level pose estimation on the REAL275, CAMERA25, and HouseCat6D benchmarks.

Abstract

Category-level object pose estimation is a crucial yet challenging task in both academia and industry, and has achieved remarkable success by leveraging keypoint-based correspondence paradigms. However, most existing methods increasingly rely on stronger feature learning while overlooking whether the established correspondences are geometrically stable across diverse perturbations. This often results in fragile pose recovery under intra-class shape variations and occlusions. To tackle this challenge, we develop a novel Triangle-Invariant Geometric Consistency Learning for Category-Level Object Pose Estimation (TriCons-Pose) to anchor stable keypoints and aggregate pose-invariant cues, yielding reliable canonical mapping and accurate pose estimation. Specifically, a Structure-Consistent Keypoint Detector (SCKD) is designed to identify robust keypoints by enforcing cross-view structural consistency via normalized pairwise distance matching. Moreover, we propose a Pose-Invariant Geometric Aggregator (PIGA) to augment keypoint representations by injecting triangle-based pose-invariant descriptors into a local-to-global attention mechanism. The proposed framework is optimized using standard objective functions while incorporating an additional geometry consistency loss. Extensive experiments on REAL275, CAMERA25, and HouseCat6D datasets demonstrate the effectiveness of the proposed approach.

Problem

• category-level object pose estimation
• geometric consistency
• keypoint correspondence
• shape variation
• occlusion

Innovation

• Triangle-Invariant
• Geometric Consistency
• Category-Level Pose Estimation
• Structure-Consistent Keypoint Detection
• Pose-Invariant Aggregation