{"paper":{"title":"BeyondMimic: From Motion Tracking to Versatile Humanoid Control via Guided Diffusion","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A compact motion-tracking setup plus classifier-guided latent diffusion lets one humanoid policy master diverse agile skills and solve unseen tasks zero-shot.","cross_cats":[],"primary_cat":"cs.RO","authors_text":"C. Karen Liu, Guy Tevet, Koushil Sreenath, Qiayuan Liao, Takara E. Truong, Xiaoyu Huang, Yuman Gao","submitted_at":"2025-08-11T17:55:26Z","abstract_excerpt":"The human-like form of humanoid robots positions them uniquely to achieve the agility and versatility in motor skills that humans possess. Learning from human demonstrations offers a scalable approach to acquiring these capabilities. However, prior works either produce unnatural motions or rely on motion-specific tuning to achieve satisfactory naturalness. Furthermore, these methods are often motion- or goal-specific, lacking the versatility to compose diverse skills, especially when solving unseen tasks. We present BeyondMimic, a framework that scales to diverse motions and carries the versat"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"A compact motion-tracking formulation enables mastering a wide range of radically agile behaviors, including aerial cartwheels, spin-kicks, flip-kicks, and sprinting, with a single setup and shared hyperparameters, while a unified latent diffusion model with classifier guidance solves downstream tasks never encountered during training and transfers zero-shot to real hardware.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That classifier guidance during diffusion sampling can reliably steer toward novel objectives (motion inpainting, teleoperation, obstacle avoidance) while preserving motion naturalness and stability without task-specific retraining or post-hoc tuning that would undermine the single-setup claim.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"BeyondMimic combines compact motion tracking with a unified guided latent diffusion model to master diverse agile behaviors from human demos and solve unseen downstream tasks via test-time classifier guidance.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A compact motion-tracking setup plus classifier-guided latent diffusion lets one humanoid policy master diverse agile skills and solve unseen tasks zero-shot.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"b1932b5fa9c0fc1b50f5696f609d8b5927742152a6a476c18feb715751c5c95b"},"source":{"id":"2508.08241","kind":"arxiv","version":4},"verdict":{"id":"ffbce39c-98cd-4932-9fbd-67455a74ae05","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T23:09:48.346184Z","strongest_claim":"A compact motion-tracking formulation enables mastering a wide range of radically agile behaviors, including aerial cartwheels, spin-kicks, flip-kicks, and sprinting, with a single setup and shared hyperparameters, while a unified latent diffusion model with classifier guidance solves downstream tasks never encountered during training and transfers zero-shot to real hardware.","one_line_summary":"BeyondMimic combines compact motion tracking with a unified guided latent diffusion model to master diverse agile behaviors from human demos and solve unseen downstream tasks via test-time classifier guidance.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That classifier guidance during diffusion sampling can reliably steer toward novel objectives (motion inpainting, teleoperation, obstacle avoidance) while preserving motion naturalness and stability without task-specific retraining or post-hoc tuning that would undermine the single-setup claim.","pith_extraction_headline":"A compact motion-tracking setup plus classifier-guided latent diffusion lets one humanoid policy master diverse agile skills and solve unseen tasks zero-shot."},"references":{"count":94,"sample":[{"doi":"","year":2016,"title":"Kuindersma,et al., Optimization-based locomotion planning, estimation, and control design for the atlas humanoid robot.Autonomous robots40(3), 429–455 (2016)","work_id":"f4657e60-3e20-474e-af85-92ffb4f9b5f9","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"P. M. Wensing,et al., Optimization-based control for dynamic legged robots.IEEE Transactions on Robotics40, 43–63 (2023)","work_id":"c03a17e1-3d02-4edc-8edb-cdd0039c033d","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2003,"title":"Kajita,et al., Biped walking pattern generation by using preview control of zero- moment point, in2003 IEEE international conference on robotics and automation (Cat","work_id":"429a04cc-fac7-49f0-9f99-3a5d8abfb43b","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2006,"title":"J. Pratt, J. Carff, S. Drakunov, A. Goswami, Capture point: A step toward humanoid push recovery, in2006 6th IEEE-RAS international conference on humanoid robots(Ieee) (2006), pp. 200–207","work_id":"a2e90583-e38a-48f1-abb0-6dbc70a1d361","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2014,"title":"R. Deits, R. Tedrake, Footstep planning on uneven terrain with mixed-integer convex opti- mization, in2014 IEEE-RAS international conference on humanoid robots(IEEE) (2014), pp. 279–286","work_id":"330d2432-6922-4a72-b969-a66fed94cfad","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":94,"snapshot_sha256":"dfc0820d0b81916122074258ae9060b0dc75a03b1338ea6b99339567b087db91","internal_anchors":3},"formal_canon":{"evidence_count":3,"snapshot_sha256":"af991a478e87cbb30324b2a16c52dc6f221141b0c4de426c267c9e2f920bdd5b"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}