Grasp-Anything: Large-scale Grasp Dataset from Foundation Models

TL;DR

This paper introduces Grasp-Anything, a large-scale, diverse grasp dataset synthesized from foundation models, enabling improved zero-shot grasp detection and robotic applications by leveraging extensive real-world knowledge.

Contribution

The paper presents a novel large-scale grasp dataset generated from foundation models, significantly increasing diversity and enabling zero-shot grasp detection in robotics.

Findings

Successfully facilitates zero-shot grasp detection

Enables real-world robotic grasping experiments

Surpasses prior datasets in size and diversity

Abstract

Foundation models such as ChatGPT have made significant strides in robotic tasks due to their universal representation of real-world domains. In this paper, we leverage foundation models to tackle grasp detection, a persistent challenge in robotics with broad industrial applications. Despite numerous grasp datasets, their object diversity remains limited compared to real-world figures. Fortunately, foundation models possess an extensive repository of real-world knowledge, including objects we encounter in our daily lives. As a consequence, a promising solution to the limited representation in previous grasp datasets is to harness the universal knowledge embedded in these foundation models. We present Grasp-Anything, a new large-scale grasp dataset synthesized from foundation models to implement this solution. Grasp-Anything excels in diversity and magnitude, boasting 1M samples with text descriptions and more than 3M objects, surpassing prior datasets. Empirically, we show that Grasp-Anything successfully facilitates zero-shot grasp detection on vision-based tasks and real-world robotic experiments. Our dataset and code are available at https://grasp-anything-2023.github.io.