Xie, ZhaomingLing, Hung YuKim, Nam HeePanne, Michiel van deBender, Jan and Popa, Tiberiu2020-10-162020-10-1620201467-8659https://doi.org/10.1111/cgf.14115https://diglib.eg.org:443/handle/10.1111/cgf14115Humans are highly adept at walking in environments with foot placement constraints, including stepping-stone scenarios where footstep locations are fully constrained. Finding good solutions to stepping-stone locomotion is a longstanding and fundamental challenge for animation and robotics. We present fully learned solutions to this difficult problem using reinforcement learning. We demonstrate the importance of a curriculum for efficient learning and evaluate four possible curriculum choices compared to a non-curriculum baseline. Results are presented for a simulated humanoid, a realistic bipedal robot simulation and a monster character, in each case producing robust, plausible motions for challenging stepping stone sequences and terrains.Computing methodologiesReinforcement learningPhysical simulation10.1111/cgf.14115213-224