FineGrasp: Towards Robust Grasping for Delicate Objects

TL;DR

FineGrasp is a new robotic grasping approach that improves handling of delicate and small objects by refining network design, label normalization, and leveraging mixed sim-to-real training, leading to better grasping success.

Contribution

The paper introduces FineGrasp, a novel grasping method with network modifications, label normalization, and a new dataset, enhancing grasping of delicate and small objects.

Findings

Significant improvement in grasping small objects

Enhanced grasp success rate for delicate regions

Effective mixed sim-to-real training strategy

Abstract

Recent advancements in robotic grasping have led to its integration as a core module in many manipulation systems. For instance, language-driven semantic segmentation enables the grasping of any designated object or object part. However, existing methods often struggle to generate feasible grasp poses for small objects or delicate components, potentially causing the entire pipeline to fail. To address this issue, we propose a novel grasping method, FineGrasp, which introduces improvements in three key aspects. First, we introduce multiple network modifications to enhance the ability of to handle delicate regions. Second, we address the issue of label imbalance and propose a refined graspness label normalization strategy. Third, we introduce a new simulated grasp dataset and show that mixed sim-to-real training further improves grasp performance. Experimental results show significant improvements, especially in grasping small objects, and confirm the effectiveness of our system in semantic grasping.