Exploring Category-level Articulated Object Pose Tracking on SE(3) Manifolds

2025-11-08

AI Summary

This work addresses the long-standing challenge of category-level pose tracking for articulated objects on the SE(3) manifold—a problem hindered by strong kinematic constraints and poor cross-category generalization. We propose PPF-Tracker, the first framework to integrate point-pair features (PPFs) with point-cloud pseudo-normalization in SE(3), enabling geometrically consistent representation. Our method introduces a semantic-driven, unified joint-axis constraint and jointly optimizes pose estimation via voting-based initialization and physics-aware kinematic modeling, thereby enforcing both geometric fidelity and physical plausibility. Extensive experiments on synthetic and real-world datasets demonstrate significant improvements in multi-frame tracking accuracy and robustness over prior methods. Notably, PPF-Tracker achieves strong cross-category generalization without category-specific training, validating its practical applicability in robotic manipulation and everyday human–object interaction tasks.

Technology Category

• Computer Vision: Motion & Tracking
• Intelligent Robots: State Estimation
• Machine Learning: Learning with Manifolds

Application Category

• Security and Privacy: Tracking, profiling, and countermeasures against them
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Articulated objects are prevalent in daily life and robotic manipulation tasks. However, compared to rigid objects, pose tracking for articulated objects remains an underexplored problem due to their inherent kinematic constraints. To address these challenges, this work proposes a novel point-pair-based pose tracking framework, termed extbf{PPF-Tracker}. The proposed framework first performs quasi-canonicalization of point clouds in the SE(3) Lie group space, and then models articulated objects using Point Pair Features (PPF) to predict pose voting parameters by leveraging the invariance properties of SE(3). Finally, semantic information of joint axes is incorporated to impose unified kinematic constraints across all parts of the articulated object. PPF-Tracker is systematically evaluated on both synthetic datasets and real-world scenarios, demonstrating strong generalization across diverse and challenging environments. Experimental results highlight the effectiveness and robustness of PPF-Tracker in multi-frame pose tracking of articulated objects. We believe this work can foster advances in robotics, embodied intelligence, and augmented reality. Codes are available at https://github.com/mengxh20/PPFTracker.

Problem

• Tracking articulated object poses with SE(3) manifold constraints
• Addressing kinematic constraints in multi-frame pose estimation
• Developing category-level generalization for diverse articulated objects

Innovation

• Quasi-canonicalization of point clouds in SE(3) space
• Using Point Pair Features for pose voting prediction
• Incorporating joint axis semantic kinematic constraints