PUGS: Zero-shot Physical Understanding with Gaussian Splatting
This addresses the problem of physical understanding for robotic systems, enabling improved real-world grasping tasks, though it appears incremental as it builds on existing Gaussian splatting techniques.
The paper tackles the challenge of predicting physical properties like mass, friction, and hardness for objects in the wild by proposing PUGS, a zero-shot method using Gaussian splatting, which achieves new state-of-the-art results on the ABO-500 mass prediction benchmark.
Current robotic systems can understand the categories and poses of objects well. But understanding physical properties like mass, friction, and hardness, in the wild, remains challenging. We propose a new method that reconstructs 3D objects using the Gaussian splatting representation and predicts various physical properties in a zero-shot manner. We propose two techniques during the reconstruction phase: a geometry-aware regularization loss function to improve the shape quality and a region-aware feature contrastive loss function to promote region affinity. Two other new techniques are designed during inference: a feature-based property propagation module and a volume integration module tailored for the Gaussian representation. Our framework is named as zero-shot physical understanding with Gaussian splatting, or PUGS. PUGS achieves new state-of-the-art results on the standard benchmark of ABO-500 mass prediction. We provide extensive quantitative ablations and qualitative visualization to demonstrate the mechanism of our designs. We show the proposed methodology can help address challenging real-world grasping tasks. Our codes, data, and models are available at https://github.com/EverNorif/PUGS