Generalization of human grasping for multi-fingered robot hands

By Heni Ben Amor, Oliver Kroemer, U. Hillenbrand, Gerhard Neumann, Jan Peters. 2012

118 citations recorded by the source corpus.

Multi-fingered robot grasping is a challenging problem that is difficult to tackle using hand-coded programs. In this paper we present an imitation learning approach for learning and generalizing grasping skills based on human demonstrations. To this end, we split the task of synthesizing a grasping motion into three parts: (1) learning efficient grasp representations from human demonstrations, (2) warping contact points onto new objects, and (3) optimizing and executing the reach-and-grasp movements. We learn low-dimensional latent grasp spaces for different grasp types, which form the basis for a novel extension to dynamic motor primitives. These latent-space dynamic motor primitives are used to synthesize entire reach-and-grasp movements. We evaluated our method on a real humanoid robot. The results of the experiment demonstrate the robustness and versatility of our approach.

Subfields: Biomedical Engineering, Cognitive Neuroscience, Control and Systems Engineering

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 Cognitive Neuroscience

In Control and Systems Engineering

Citations in the corpus

Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations (2017, background): “expense of reduced flexibility, or on acquiring controllers for relatively simple behaviors such as grasping [2] or rolling objects in the fingers [52], often with low degree of freedom manipulators.”
Language-Conditioned Imitation Learning for Robot Manipulation Tasks (2020, background): “Learning robot control policies by imitation [31] is an appealing approach to skill acquisition and has been successfully applied to several tasks, including locomotion, grasping, and even table tennis [8, 2, 25].”
Learning Dexterous Grasping with Object-Centric Visual Affordances (2020, methodology): “explores related open-loop strategies with complex controllers, but, unlike our method, assumes access to the full 3D model of the objects [10], [2], [4], [7], [44].”
Grasping in the Wild: Learning 6DoF Closed-Loop Grasping From Low-Cost Demonstrations (2019, background): “On the other hand, recording videos of direct interactions between human hands and objects does not require expert knowledge from the subject [27], [28].”
One-shot learning and generation of dexterous grasps for novel objects (2016, methodology): “Another class assumes the existence of a full 3D model of the object’s shape (Ben Amor et al., 2012; Hillenbrand and Roa, 2012).”
Active vision for dexterous grasping of novel objects (2016, background): “Grasping of novel objects is a hard problem on which there has been steady progress [12], [13], [10], [17], [9], [20], [8], [4], [18], [5].”
A Review of Robot Learning for Manipulation: Challenges, Representations, and Algorithms (2019, background): “, the tip of a tool (Edsinger and Kemp, 2006)), or use non-rigid registration or geometry warping to determine correspondences of entire sets of points across task instances (Schulman et al., 2013; Hillenbrand and Roa, 2012; Stückler and Behnke, 2014; Rodriguez and Behnke, 2018; Amor et al., 2012).”
ContactGrasp: Functional Multi-finger Grasp Synthesis from Contact (2019, background): “Ben Amor et al [31] learn a low-dimensional grasp space from human demonstrations acquired using a dataglove.”