PUGS: Zero-shot Physical Understanding with Gaussian Splatting

TL;DR

PUGS introduces a novel zero-shot framework using Gaussian splatting for understanding physical properties of objects, enhancing robotic perception and grasping in real-world scenarios.

Contribution

The paper presents a new zero-shot method for predicting physical properties of 3D objects using Gaussian splatting, with novel regularization and inference techniques.

Findings

Achieves state-of-the-art mass prediction on ABO-500 benchmark.

Demonstrates effectiveness in real-world grasping tasks.

Provides extensive ablation studies and visualizations.

Abstract

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