{"paper":{"title":"Learning Human-Intention Priors from Large-Scale Human Demonstrations for Robotic Manipulation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"MoT-HRA learns human-intention priors from 2.2 million video episodes to guide more reliable robot manipulation.","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Guangyu Chen, Jinkun Liu, Wenbo Ding, Yifan Xie, Yuan Wang, Yu Sun","submitted_at":"2026-04-27T16:42:18Z","abstract_excerpt":"Human videos contain rich manipulation priors, but using them for robot learning remains difficult because raw observations entangle scene understanding, human motion, and embodiment-specific action. We introduce MoT-HRA, a hierarchical vision-language-action framework that learns human-intention priors from large-scale human demonstrations. We first curate HA-2.2M, a 2.2M-episode action-language dataset reconstructed from heterogeneous human videos through hand-centric filtering, spatial reconstruction, temporal segmentation, and language alignment. On top of this dataset, MoT-HRA factorizes "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Experiments on hand motion generation, simulated manipulation, and real-world robot tasks show that MoT-HRA improves motion plausibility and robust control under distribution shift.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The hand-centric filtering, spatial reconstruction, temporal segmentation, and language alignment steps used to curate HA-2.2M from heterogeneous human videos successfully extract embodiment-agnostic human-intention priors without introducing substantial artifacts, biases, or loss of critical information.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"MoT-HRA learns embodiment-agnostic human-intention priors from the HA-2.2M dataset of 2.2M human video episodes through a three-expert hierarchy to improve robotic motion plausibility and robustness under distribution shift.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"MoT-HRA learns human-intention priors from 2.2 million video episodes to guide more reliable robot manipulation.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"66c603715eb6d7cdf74b6843da26f8261679f25ef47635e1e19d0f3a673bd240"},"source":{"id":"2604.24681","kind":"arxiv","version":2},"verdict":{"id":"5ab851d7-5b8c-47f4-90b0-e0462b72c8ef","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-08T02:45:55.774426Z","strongest_claim":"Experiments on hand motion generation, simulated manipulation, and real-world robot tasks show that MoT-HRA improves motion plausibility and robust control under distribution shift.","one_line_summary":"MoT-HRA learns embodiment-agnostic human-intention priors from the HA-2.2M dataset of 2.2M human video episodes through a three-expert hierarchy to improve robotic motion plausibility and robustness under distribution shift.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The hand-centric filtering, spatial reconstruction, temporal segmentation, and language alignment steps used to curate HA-2.2M from heterogeneous human videos successfully extract embodiment-agnostic human-intention priors without introducing substantial artifacts, biases, or loss of critical information.","pith_extraction_headline":"MoT-HRA learns human-intention priors from 2.2 million video episodes to guide more reliable robot manipulation."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.24681/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-21T06:35:50.393016Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T21:52:19.836490Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"272f2bb85cf96cb87582202480ecbbd3958db1baffd10a154ca3ea0257545eec"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}