Saved in:
Bibliographic Details
Main Author: Guo, Dingkun
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2312.15346
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913185948237824
author Guo, Dingkun
author_facet Guo, Dingkun
contents Learning from human demonstrations has exhibited remarkable achievements in robot manipulation. However, the challenge remains to develop a robot system that matches human capabilities and data efficiency in learning and generalizability, particularly in complex, unstructured real-world scenarios. We propose a system that processes RGBD videos to translate human actions to robot primitives and identifies task-relevant key poses of objects using Grounded Segment Anything. We then address challenges for robots in replicating human actions, considering the human-robot differences in kinematics and collision geometry. To test the effectiveness of our system, we conducted experiments focusing on manual dishwashing. With a single human demonstration recorded in a mockup kitchen, the system achieved 50-100% success for each step and up to a 40% success rate for the whole task with different objects in a home kitchen. Videos are available at https://robot-dishwashing.github.io
format Preprint
id arxiv_https___arxiv_org_abs_2312_15346
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Learning Multi-Step Manipulation Tasks from A Single Human Demonstration
Guo, Dingkun
Robotics
Learning from human demonstrations has exhibited remarkable achievements in robot manipulation. However, the challenge remains to develop a robot system that matches human capabilities and data efficiency in learning and generalizability, particularly in complex, unstructured real-world scenarios. We propose a system that processes RGBD videos to translate human actions to robot primitives and identifies task-relevant key poses of objects using Grounded Segment Anything. We then address challenges for robots in replicating human actions, considering the human-robot differences in kinematics and collision geometry. To test the effectiveness of our system, we conducted experiments focusing on manual dishwashing. With a single human demonstration recorded in a mockup kitchen, the system achieved 50-100% success for each step and up to a 40% success rate for the whole task with different objects in a home kitchen. Videos are available at https://robot-dishwashing.github.io
title Learning Multi-Step Manipulation Tasks from A Single Human Demonstration
topic Robotics
url https://arxiv.org/abs/2312.15346