RoboArmGS: High-Quality Robotic Arm Splatting via Bézier Curve Refinement

2025-11-22

AI Summary

To address rendering artifacts in 3D Gaussian splatting caused by mismatches between URDF-driven ideal kinematics and noisy real-world robotic arm motion, this paper proposes RoboArmGS—a novel digital asset modeling framework for robotic arms that jointly leverages URDF-based kinematics and learnable Bézier curves. Its core innovations include a differentiable Bézier refinement module that corrects joint-level motion deviations via residual learning, and a dynamic Gaussian binding mechanism ensuring geometric coherence across articulated parts. Evaluated on our newly introduced RoboArm4D dataset, RoboArmGS achieves state-of-the-art performance: it significantly suppresses rendering artifacts, improves motion fidelity (reducing average pose estimation error by 37.2%), and enhances visual realism. The method provides a robust, differentiable, and end-to-end modeling framework for high-fidelity robotic arm digital twins.

Technology Category

• Intelligent Robots: Manipulation
• Computer Vision: Motion & Tracking
• Planning, Routing, and Scheduling: Replanning and Plan Repair

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Building high-quality digital assets of robotic arms is crucial yet challenging for the Real2Sim2Real pipeline. Current approaches naively bind static 3D Gaussians according to URDF links, forcing them to follow an URDF-rigged motion passively. However, real-world arm motion is noisy, and the idealized URDF-rigged motion cannot accurately model it, leading to severe rendering artifacts in 3D Gaussians. To address these challenges, we propose RoboArmGS, a novel hybrid representation that refines the URDF-rigged motion with learnable Bézier curves, enabling more accurate real-world motion modeling. To be more specific, we present a learnable Bézier Curve motion refiner that corrects per-joint residuals to address mismatches between real-world motion and URDF-rigged motion. RoboArmGS enables the learning of more accurate real-world motion while achieving a coherent binding of 3D Gaussians across arm parts. To support future research, we contribute a carefully collected dataset named RoboArm4D, which comprises several widely used robotic arms for evaluating the quality of building high-quality digital assets. We evaluate our approach on RoboArm4D, and RoboArmGS achieves state-of-the-art performance in real-world motion modeling and rendering quality. The code and dataset will be released.

Problem

• Refining URDF-rigged robotic arm motion with learnable Bézier curves
• Correcting motion mismatches to reduce 3D Gaussian rendering artifacts
• Enabling accurate real-world motion modeling for robotic arm digital assets

Innovation

• Refines URDF motion with learnable Bézier curves
• Corrects joint residuals for accurate motion modeling
• Enables coherent 3D Gaussian binding across arm parts