Adversarial Reinforcement Learning for Procedural Content Generation

Training RL agents to solve novel environments is a notoriously difficult task. Here we present a new approach ARLPCG: Adversarial Reinforcement Learning for Procedural Content Generation, which procedurally generates and tests previously unseen environments with an auxiliary input as a control variable. The procedurally generated environments induces state diversity which increases the generalizability of the trained agents. ARLPCG deploys an adversarial model with one PCG RL agent (called Generator) and one solving RL agent (called Solver). The Generator receives a reward signal based on the Solver's performance, which encourages the environment design to be challenging but not impossible. To further drive diversity and control of the environment generation, we propose using auxiliary inputs for the Generator. The benefit is two-fold: Firstly, the Solver achieves better generalization through the Generator's generated challenges. Secondly, the trained Generator can be used as a creator of novel environments that, together with the Solver, can be shown to be solvable. We create two types of 3D environments to validate our model, representing two popular game genres: a third-person platformer and a racing game. In these cases, we show that ARLPCG has a significantly better solve ratio, and that the auxiliary inputs renders the levels creation controllable to a certain degree. For a video compilation of the results please visit https://youtu.be/z7q2PtVsT0I.