Mask Atari for Deep Reinforcement Learning as POMDP Benchmarks
This provides a new benchmark for researchers working on POMDP problems in deep reinforcement learning, though it is incremental as it builds on existing Atari environments.
The authors tackled the lack of a benchmark for partially observable Markov decision processes (POMDPs) in deep reinforcement learning by introducing Mask Atari, a simulation environment based on Atari 2600 games with controllable masks, and evaluated several baselines to demonstrate its utility.
We present Mask Atari, a new benchmark to help solve partially observable Markov decision process (POMDP) problems with Deep Reinforcement Learning (DRL)-based approaches. To achieve a simulation environment for the POMDP problems, Mask Atari is constructed based on Atari 2600 games with controllable, moveable, and learnable masks as the observation area for the target agent, especially with the active information gathering (AIG) setting in POMDPs. Given that one does not yet exist, Mask Atari provides a challenging, efficient benchmark for evaluating the methods that focus on the above problem. Moreover, the mask operation is a trial for introducing the receptive field in the human vision system into a simulation environment for an agent, which means the evaluations are not biased from the sensing ability and purely focus on the cognitive performance of the methods when compared with the human baseline. We describe the challenges and features of our benchmark and evaluate several baselines with Mask Atari.