SparseGrasp: Robotic Grasping via 3D Semantic Gaussian Splatting from Sparse Multi-View RGB Images

TL;DR

SparseGrasp is a new robotic grasping system that efficiently uses sparse multi-view RGB images, incorporates semantic understanding, and rapidly updates scenes for better performance in dynamic environments.

Contribution

It introduces a novel approach combining 3D Gaussian Splatting, semantic awareness, and PCA-based feature compression for fast, scene-adaptive robotic grasping from sparse views.

Findings

Outperforms state-of-the-art in speed and adaptability

Enables multi-turn grasping in changeable environments

Maintains high fidelity with sparse supervision

Abstract

Language-guided robotic grasping is a rapidly advancing field where robots are instructed using human language to grasp specific objects. However, existing methods often depend on dense camera views and struggle to quickly update scenes, limiting their effectiveness in changeable environments. In contrast, we propose SparseGrasp, a novel open-vocabulary robotic grasping system that operates efficiently with sparse-view RGB images and handles scene updates fastly. Our system builds upon and significantly enhances existing computer vision modules in robotic learning. Specifically, SparseGrasp utilizes DUSt3R to generate a dense point cloud as the initialization for 3D Gaussian Splatting (3DGS), maintaining high fidelity even under sparse supervision. Importantly, SparseGrasp incorporates semantic awareness from recent vision foundation models. To further improve processing efficiency, we repurpose Principal Component Analysis (PCA) to compress features from 2D models. Additionally, we introduce a novel render-and-compare strategy that ensures rapid scene updates, enabling multi-turn grasping in changeable environments. Experimental results show that SparseGrasp significantly outperforms state-of-the-art methods in terms of both speed and adaptability, providing a robust solution for multi-turn grasping in changeable environment.