Abstract:This article investigates the challenge of achieving functional tool-use grasping with high-DoF anthropomorphic hands, with the aim of enabling anthropomorphic hands to perform tasks that require human-like manipulation and tool-use. However, accomplishing human-like grasping in real robots present many challenges, including obtaining diverse functional grasps for a wide variety of objects, handling generalization ability for kinematically diverse robot hands and precisely completing object shapes from a single-view perception. To tackle these challenges, we propose a six-step grasp synthesis algorithm based on fine-grained contact modeling that generates physically plausible and human-like functional grasps for category-level objects with minimal human demonstrations. With the contact-based optimization and learned dense shape correspondence, the proposed algorithm is adaptable to various objects in same category and a board range of robot hand models. To further demonstrate the robustness of the framework, over 10K functional grasps are synthesized to train our neural network, named DexFG-Net, which generates diverse sets of human-like functional grasps based on the reconstructed object model produced by a shape completion module. The proposed framework is extensively validated in simulation and on a real robot platform. Simulation experiments demonstrate that our method outperforms baseline methods by a large margin in terms of grasp functionality and success rate. Real robot experiments show that our method achieved an overall success rate of 79\% and 68\% for tool-use grasp on 3-D printed and real test objects, respectively, using a 5-Finger Schunk Hand. The experimental results indicate a step towards human-like grasping with anthropomorphic hands.

Learning Realistic and Reasonable Grasps for Anthropomorphic Hand in Cluttered Scenes

Learning Object Affordance with Contact and Grasp Generation

A Cascaded Deep Learning Framework for Real-time and Robust Grasp Planning

Realistic Virtual Hand Modeling with Applications for Virtual Grasping

Efficient and Robust Robotic Grasping in Cluttered Scenes: A Point Cloud-Based Approach with Heuristic Evaluation.

Generalized Anthropomorphic Functional Grasping with Minimal Demonstrations

HGC-Net: Deep Anthropomorphic Hand Grasping in Clutter

D-Grasp: Physically Plausible Dynamic Grasp Synthesis for Hand-Object Interactions

FunctionalGrasp: Learning Functional Grasp for Robots Via Semantic Hand-Object Representation

Plausible and Diverse Human Hand Grasping Motion Generation

Learning Robust Real-World Dexterous Grasping Policies via Implicit Shape Augmentation

Single-View Scene Point Cloud Human Grasp Generation

Learning Human-to-Robot Dexterous Handovers for Anthropomorphic Hand

Learning Human-Like Functional Grasping for Multifinger Hands From Few Demonstrations

Anthropomorphic Grasping with Neural Object Shape Completion

Multi-fingered Robotic Hand Grasping in Cluttered Environments through Hand-object Contact Semantic Mapping

Toward Human-Like Grasp: Dexterous Grasping via Semantic Representation of Object-Hand

Learning Accurate and Efficient Three-Finger Grasp Generation in Clutters with an Auto-Annotated Large-Scale Dataset

DexGraspNet 2.0: Learning Generative Dexterous Grasping in Large-scale Synthetic Cluttered Scenes

Reasonable Grasping Based on Hierarchical Decomposition Models of Unknown Objects.

MVGrasp: Real-time multi-view 3D object grasping in highly cluttered environments