Improved GQ-CNN: Deep Learning Model for Planning Robust Grasps

TL;DR

This paper enhances the GQ-CNN model for robotic grasping by introducing a new architecture and practical improvements, significantly increasing validation accuracy and improving success rates with unknown objects.

Contribution

The paper presents a novel architecture for GQ-CNN and practical modifications that boost validation accuracy from 92.2% to 95.8% and 85.9% to 88.0%.

Findings

Validation accuracy improved from 92.2% to 95.8%.

Object-wise validation accuracy increased from 85.9% to 88.0%.

Enhanced model achieves higher grasp success rates on unseen objects.

Abstract

Recent developments in the field of robot grasping have shown great improvements in the grasp success rates when dealing with unknown objects. In this work we improve on one of the most promising approaches, the Grasp Quality Convolutional Neural Network (GQ-CNN) trained on the DexNet 2.0 dataset. We propose a new architecture for the GQ-CNN and describe practical improvements that increase the model validation accuracy from 92.2% to 95.8% and from 85.9% to 88.0% on respectively image-wise and object-wise training and validation splits.