Published on:
12 March 2024
Primary Category:
Robotics
Paper Authors:
Aadhithya Iyer,
Zhuoran Peng,
Yinlong Dai,
Irmak Guzey,
Siddhant Haldar,
Soumith Chintala,
Lerrel Pinto
Presents OPEN TEACH, an open-source unified robot teleoperation system using affordable VR
Supports real-time control of various robots like multi-fingered hands and bimanual arms
Allows robot manipulation through natural hand gestures at up to 90Hz with smooth feedback
Demonstrates versatility across 38 tasks on different robots
Enables policy learning for 10 dexterous manipulation tasks
Teleoperation system for robot control
The paper presents OPEN TEACH, an open-source teleoperation system using affordable VR headsets for intuitive control of various robots like multi-fingered hands and bimanual arms. It allows real-time robot manipulation at up to 90Hz through natural hand gestures with smooth visual feedback. Experiments across 38 tasks on different robots demonstrate versatility. A user study shows significant improvement over prior systems. The collected data enables policy learning for 10 dexterous manipulation tasks.
Learning dexterity from human demonstrations
Teleoperation with haptic feedback using low-cost tactile sensors
Simplifying Medical Training with Natural Gestures
Robot communication through mixed reality
Teleoperation system for mobile robot control
Robotic system learns new manipulation skills from user demonstrations
No comments yet, be the first to start the conversation...
Sign up to comment on this paper