Selectively rendering visuals in subsets of camera groups #833
-
|
I'm wondering if there is any way to create a purely visual entity (no collision geometry) that gets rendered only in certain camera subgroups such as the viewer/human cameras but not the sensor cameras. I am hoping to visualize the target pose of the PD EE Pose controller to help me with teleoperation, but don't want it to appear in the sensor observations. |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
|
Yes, see PickCube task as an example. You can register some objects as being part of the hidden objects list. Anytime you render sensors it will be hidden. The moment the human render is called it will be unhidden and for objects if you don't add a collision mesh then its visual only |
Beta Was this translation helpful? Give feedback.
-
|
Thanks! Literally just found the hidden_objects argument now. |
Beta Was this translation helpful? Give feedback.
Yes, see PickCube task as an example. You can register some objects as being part of the hidden objects list. Anytime you render sensors it will be hidden. The moment the human render is called it will be unhidden
and for objects if you don't add a collision mesh then its visual only