I'm trying to create a camera in the scene and then get the segmentation mask from it. How would I do this?
Context for my specific use case: This is after env.launch() is called, I do self.cam = VisionSensor.create([img_size, img_size])
and it works, but how do I get the mask from this? For the cameras given in the scene there exists corresponding mask cams, i.e., for cameraovershoulderleft there is cameraovershouldlerft_mask, but i check the ttm and there isn't any cameraovershoulder_mask in the scene.
Hi,
I'm trying to create a camera in the scene and then get the segmentation mask from it. How would I do this?
Context for my specific use case: This is after env.launch() is called, I do
self.cam = VisionSensor.create([img_size, img_size])
and it works, but how do I get the mask from this? For the cameras given in the scene there exists corresponding mask cams, i.e., for cameraovershoulderleft there is cameraovershouldlerft_mask, but i check the ttm and there isn't any cameraovershoulder_mask in the scene.looking at this commit https://github.com/stepjam/RLBench/commit/d0006faeb195b52a6c19306222c05d7cc40f35b9 maybe the mask cams are done manually? How can I dynamically make a cam have a corresponding mask one?