ALLSTEPS: Curriculum-driven Learning of Stepping Stone Skills

dc.contributor.authorXie, Zhaomingen_US
dc.contributor.authorLing, Hung Yuen_US
dc.contributor.authorKim, Nam Heeen_US
dc.contributor.authorPanne, Michiel van deen_US
dc.contributor.editorBender, Jan and Popa, Tiberiuen_US
dc.date.accessioned2020-10-16T06:25:57Z
dc.date.available2020-10-16T06:25:57Z
dc.date.issued2020
dc.description.abstractHumans 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.en_US
dc.description.number8
dc.description.sectionheadersCharacter Animation 1
dc.description.seriesinformationComputer Graphics Forum
dc.description.volume39
dc.identifier.doi10.1111/cgf.14115
dc.identifier.issn1467-8659
dc.identifier.pages213-224
dc.identifier.urihttps://doi.org/10.1111/cgf.14115
dc.identifier.urihttps://diglib.eg.org:443/handle/10.1111/cgf14115
dc.publisherThe Eurographics Association and John Wiley & Sons Ltd.en_US
dc.subjectComputing methodologies
dc.subjectReinforcement learning
dc.subjectPhysical simulation
dc.titleALLSTEPS: Curriculum-driven Learning of Stepping Stone Skillsen_US
Files
Collections