Simultaneous Semantic and Collision Learning for 6-DoF Grasp Pose Estimation
This work addresses the problem of enabling robots to grasp objects in cluttered environments more effectively, representing an incremental advance by integrating multiple tasks into a unified framework.
The paper tackles the challenge of 6-DoF grasp pose estimation in cluttered scenes by proposing a simultaneous multi-task learning framework that jointly predicts grasp poses, semantic segmentation, and collision information, achieving a +4.08 AP improvement over prior state-of-the-art methods on a public dataset and demonstrating high success rates in real robot grasping.
Grasping in cluttered scenes has always been a great challenge for robots, due to the requirement of the ability to well understand the scene and object information. Previous works usually assume that the geometry information of the objects is available, or utilize a step-wise, multi-stage strategy to predict the feasible 6-DoF grasp poses. In this work, we propose to formalize the 6-DoF grasp pose estimation as a simultaneous multi-task learning problem. In a unified framework, we jointly predict the feasible 6-DoF grasp poses, instance semantic segmentation, and collision information. The whole framework is jointly optimized and end-to-end differentiable. Our model is evaluated on large-scale benchmarks as well as the real robot system. On the public dataset, our method outperforms prior state-of-the-art methods by a large margin (+4.08 AP). We also demonstrate the implementation of our model on a real robotic platform and show that the robot can accurately grasp target objects in cluttered scenarios with a high success rate. Project link: https://openbyterobotics.github.io/sscl