Hi there! Could you tell me how can I get the camera rendered state from the point of view of agent’s eyes (as visual sensory data)? I would like to obtain agent's eyes view as black and white image ~3 times/second. So it won't be on every simulation step. Is that possible? Thanks for your help!