[Question] Extracting Segmentation mask from Gym Enviorment

[hamzmu]

I am new to Gym, is it possible to get segmentation mask from a gym environment? The only render modes I see are 'human' and 'rgb_array'. Since the engine is based on mujoco which supports getting segmentation mask, is there any way to extract the segmentation mask on an enviorment?

[Kallinteris-Andreas]

render_mode=rgb_array gives you the images with env.render() on every step, if you want to segment them you will have to do that yourself