OpenTeleVision / TeleVision

[CoRL 2024] Open-TeleVision: Teleoperation with Immersive Active Visual Feedback
https://robot-tv.github.io/
Other
651 stars 66 forks source link

From VisionProStreamer to Teaching data #21

Open zhihuiwis opened 4 months ago

zhihuiwis commented 4 months ago

How to transform from the key point detection of the hand to the action vector required by the teaching data (especially GR-1), and how is state calculating?thank you for your great work!