{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:4665DF5M5S5EBYPYPOGV6J5E5Q","short_pith_number":"pith:4665DF5M","schema_version":"1.0","canonical_sha256":"e7bdd197acecba40e1f87b8d5f27a4ec18f370293d49f26dce52bd6f60a5e901","source":{"kind":"arxiv","id":"2411.04919","version":2},"attestation_state":"computed","paper":{"title":"Stem-OB: Generalizable Visual Imitation Learning with Stem-Like Convergent Observation through Diffusion Inversion","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.RO","authors_text":"Huazhe Xu, Kaizhe Hu, Pu Hua, Yao He, Yuyao Liu, Zihang Rui","submitted_at":"2024-11-07T17:56:16Z","abstract_excerpt":"Visual imitation learning methods demonstrate strong performance, yet they lack generalization when faced with visual input perturbations, including variations in lighting and textures, impeding their real-world application. We propose Stem-OB that utilizes pretrained image diffusion models to suppress low-level visual differences while maintaining high-level scene structures. This image inversion process is akin to transforming the observation into a shared representation, from which other observations stem, with extraneous details removed. Stem-OB contrasts with data-augmentation approaches "},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2411.04919","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2024-11-07T17:56:16Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"d291adb2f1a62ce48156c92053d24467f536efaf8840df299cfa1d7f9fde84f1","abstract_canon_sha256":"a3e78f17000a8bba04a9e94ade2ffe1c45b8053c4ade02f084d7ff2020394798"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:34:39.622369Z","signature_b64":"U0mhACjpN+bLmkHlAJTr/LL6gqjS3wUYPplkUe8kdTpzO5fR8eoo7yGXxjPcQO4m3rd41+0KlNNmqLXDdJaRBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e7bdd197acecba40e1f87b8d5f27a4ec18f370293d49f26dce52bd6f60a5e901","last_reissued_at":"2026-07-05T09:34:39.621876Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:34:39.621876Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Stem-OB: Generalizable Visual Imitation Learning with Stem-Like Convergent Observation through Diffusion Inversion","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.RO","authors_text":"Huazhe Xu, Kaizhe Hu, Pu Hua, Yao He, Yuyao Liu, Zihang Rui","submitted_at":"2024-11-07T17:56:16Z","abstract_excerpt":"Visual imitation learning methods demonstrate strong performance, yet they lack generalization when faced with visual input perturbations, including variations in lighting and textures, impeding their real-world application. We propose Stem-OB that utilizes pretrained image diffusion models to suppress low-level visual differences while maintaining high-level scene structures. This image inversion process is akin to transforming the observation into a shared representation, from which other observations stem, with extraneous details removed. Stem-OB contrasts with data-augmentation approaches "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.04919","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2411.04919/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2411.04919","created_at":"2026-07-05T09:34:39.621933+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.04919v2","created_at":"2026-07-05T09:34:39.621933+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.04919","created_at":"2026-07-05T09:34:39.621933+00:00"},{"alias_kind":"pith_short_12","alias_value":"4665DF5M5S5E","created_at":"2026-07-05T09:34:39.621933+00:00"},{"alias_kind":"pith_short_16","alias_value":"4665DF5M5S5EBYPY","created_at":"2026-07-05T09:34:39.621933+00:00"},{"alias_kind":"pith_short_8","alias_value":"4665DF5M","created_at":"2026-07-05T09:34:39.621933+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.10579","citing_title":"AffordGen: Generating Diverse Demonstrations for Generalizable Object Manipulation with Afford Correspondence","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04502","citing_title":"Veo-Act: How Far Can Frontier Video Models Advance Generalizable Robot Manipulation?","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4665DF5M5S5EBYPYPOGV6J5E5Q","json":"https://pith.science/pith/4665DF5M5S5EBYPYPOGV6J5E5Q.json","graph_json":"https://pith.science/api/pith-number/4665DF5M5S5EBYPYPOGV6J5E5Q/graph.json","events_json":"https://pith.science/api/pith-number/4665DF5M5S5EBYPYPOGV6J5E5Q/events.json","paper":"https://pith.science/paper/4665DF5M"},"agent_actions":{"view_html":"https://pith.science/pith/4665DF5M5S5EBYPYPOGV6J5E5Q","download_json":"https://pith.science/pith/4665DF5M5S5EBYPYPOGV6J5E5Q.json","view_paper":"https://pith.science/paper/4665DF5M","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.04919&json=true","fetch_graph":"https://pith.science/api/pith-number/4665DF5M5S5EBYPYPOGV6J5E5Q/graph.json","fetch_events":"https://pith.science/api/pith-number/4665DF5M5S5EBYPYPOGV6J5E5Q/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4665DF5M5S5EBYPYPOGV6J5E5Q/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4665DF5M5S5EBYPYPOGV6J5E5Q/action/storage_attestation","attest_author":"https://pith.science/pith/4665DF5M5S5EBYPYPOGV6J5E5Q/action/author_attestation","sign_citation":"https://pith.science/pith/4665DF5M5S5EBYPYPOGV6J5E5Q/action/citation_signature","submit_replication":"https://pith.science/pith/4665DF5M5S5EBYPYPOGV6J5E5Q/action/replication_record"}},"created_at":"2026-07-05T09:34:39.621933+00:00","updated_at":"2026-07-05T09:34:39.621933+00:00"}