Can Observation type being extended?

[GeorgeWinsdor]

Hi, thank you for the great project!

I'm currently working on a multi-drone reinforcement learning task and I wonder if it is possible to use the KIN and RGB observation type together?

[JacopoPan]

Hello @GeorgeWinsdor

you would need to modify the observation space defined here:

gym-pybullet-drones/gym_pybullet_drones/envs/BaseRLAviary.py

Lines 243 to 256 in 2d8325d

	def _observationSpace(self):
	"""Returns the observation space of the environment.
	Returns
	-------
	ndarray
	A Box() of shape (NUM_DRONES,H,W,4) or (NUM_DRONES,12) depending on the observation type.
	"""
	if self.OBS_TYPE == ObservationType.RGB:
	return spaces.Box(low=0,
	high=255,
	shape=(self.NUM_DRONES, self.IMG_RES[1], self.IMG_RES[0], 4), dtype=np.uint8)
	elif self.OBS_TYPE == ObservationType.KIN:

But note that doing RGB for multiple drones can be really sample inefficient, if that's you goal you should probably reduce either the control/RL stepping frequency or buffer images sporadically.