Contact-GraspNet: Efficient 6-DoF Grasp Generation in Cluttered Scenes

By Martin Sundermeyer, Arsalan Mousavian, Rudolph Triebel, Dieter Fox. IEEE International Conference on Robotics and Automation, 2021

521 citations recorded by the source corpus.

Grasping unseen objects in unconstrained, cluttered environments is an essential skill for autonomous robotic manipulation. Despite recent progress in full 6-DoF grasp learning, existing approaches often consist of complex sequential pipelines that possess several potential failure points and run-times unsuitable for closed-loop grasping. Therefore, we propose an end-to-end network that efficiently generates a distribution of 6-DoF parallel-jaw grasps directly from a depth recording of a scene. Our novel grasp representation treats 3D points of the recorded point cloud as potential grasp contacts. By rooting the full 6-DoF grasp pose and width in the observed point cloud, we can reduce the dimensionality of our grasp representation to 4-DoF which greatly facilitates the learning process. Our class-agnostic approach is trained on 17 million simulated grasps and generalizes well to real world sensor data. In a robotic grasping study of unseen objects in structured clutter we achieve over 90% success rate, cutting the failure rate in half compared to a recent state-of-the-art method. Video of the real world experiments and code are available at https://research.nvidia.com/publication/2021-03_Contact-GraspNet%3A--Efficient.

Subfields: Biomedical Engineering, Control and Systems Engineering, Human-Computer Interaction

View Paper: https://www.semanticscholar.org/paper/d85d16b003955c6996fafacea7f3c075c531225f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

In Control and Systems Engineering

In Human-Computer Interaction

Citations in the corpus

Grasp'D: Differentiable Contact-rich Grasp Synthesis for Multi-fingered Hands (2022, unknown): “For parallel-jaw grippers, most recent works [47, 61, 66, 79] learn from simulation-based datasets (e.”
ManiSkill2: A Unified Benchmark for Generalizable Manipulation Skills (2023, methodology): “Notably, on PickSingleYCB , the success rate is even higher than Contact-GraspNet with motion planning.”
AnyGrasp: Robust and Efficient Grasp Perception in Spatial and Temporal Domains (2022, methodology): “Similar end-to-end networks were parallelly developed in [28], [29], and [30].”
ProgPrompt: Generating Situated Robot Task Plans using Large Language Models (2022, methodology): “We use the system of [37] to implement the policy, and use MPPI for motion generation, SceneCollisionNet [37] to avoid collisions, and generate grasp poses with Contact-GraspNet [38].”
A Survey on Vision--Language--Action Models for Embodied AI (2024, unknown): “Because the majority of current MLLMs deal with images as visual inputs, they require architectural changes to incorporate 3D visual inputs, and therefore they are usually end-to-end models.”