{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XZ7GRVPQIBJOJPG3DLGAU6RYJB","short_pith_number":"pith:XZ7GRVPQ","schema_version":"1.0","canonical_sha256":"be7e68d5f04052e4bcdb1acc0a7a384849caea639c53e74f781d3e56f292900f","source":{"kind":"arxiv","id":"2503.11646","version":1},"attestation_state":"computed","paper":{"title":"Adversarial Data Collection: Human-Collaborative Perturbations for Efficient and Robust Robotic Imitation Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Guanghui Ren, Hongsheng Li, Maoqing Yao, Shu Jiang, Si Liu, Siyuan Feng, Siyuan Huang, Yue Liao","submitted_at":"2025-03-14T17:59:07Z","abstract_excerpt":"The pursuit of data efficiency, where quality outweighs quantity, has emerged as a cornerstone in robotic manipulation, especially given the high costs associated with real-world data collection. We propose that maximizing the informational density of individual demonstrations can dramatically reduce reliance on large-scale datasets while improving task performance. To this end, we introduce Adversarial Data Collection, a Human-in-the-Loop (HiL) framework that redefines robotic data acquisition through real-time, bidirectional human-environment interactions. Unlike conventional pipelines that "},"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":"2503.11646","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-03-14T17:59:07Z","cross_cats_sorted":[],"title_canon_sha256":"5aa9291f351842c02580deb56bc6722cfb56cae21b8ef2bbaca2a47013223c4f","abstract_canon_sha256":"e509c1836f210801433f5bc60d23199a78432472e8713c328067d713078cc079"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:31:36.610339Z","signature_b64":"XuqvJXoIa+V/QVJ87inV1UB1zOEAuzV6ziJbZnwfaZp9GIXkK2qwyBDgpFY6kRREtLregZZ/5LcymtQxrq8JDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"be7e68d5f04052e4bcdb1acc0a7a384849caea639c53e74f781d3e56f292900f","last_reissued_at":"2026-07-05T10:31:36.609352Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:31:36.609352Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Adversarial Data Collection: Human-Collaborative Perturbations for Efficient and Robust Robotic Imitation Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Guanghui Ren, Hongsheng Li, Maoqing Yao, Shu Jiang, Si Liu, Siyuan Feng, Siyuan Huang, Yue Liao","submitted_at":"2025-03-14T17:59:07Z","abstract_excerpt":"The pursuit of data efficiency, where quality outweighs quantity, has emerged as a cornerstone in robotic manipulation, especially given the high costs associated with real-world data collection. We propose that maximizing the informational density of individual demonstrations can dramatically reduce reliance on large-scale datasets while improving task performance. To this end, we introduce Adversarial Data Collection, a Human-in-the-Loop (HiL) framework that redefines robotic data acquisition through real-time, bidirectional human-environment interactions. Unlike conventional pipelines that "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.11646","kind":"arxiv","version":1},"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/2503.11646/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":"2503.11646","created_at":"2026-07-05T10:31:36.609494+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.11646v1","created_at":"2026-07-05T10:31:36.609494+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.11646","created_at":"2026-07-05T10:31:36.609494+00:00"},{"alias_kind":"pith_short_12","alias_value":"XZ7GRVPQIBJO","created_at":"2026-07-05T10:31:36.609494+00:00"},{"alias_kind":"pith_short_16","alias_value":"XZ7GRVPQIBJOJPG3","created_at":"2026-07-05T10:31:36.609494+00:00"},{"alias_kind":"pith_short_8","alias_value":"XZ7GRVPQ","created_at":"2026-07-05T10:31:36.609494+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20871","citing_title":"Geometric Entropy: When Trajectory Diversity Helps and Hurts in Imitation Learning","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2607.02322","citing_title":"The Moving Eye: Enhancing VLA Spatial Generalization via Hybrid Dynamic Data Collection","ref_index":24,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XZ7GRVPQIBJOJPG3DLGAU6RYJB","json":"https://pith.science/pith/XZ7GRVPQIBJOJPG3DLGAU6RYJB.json","graph_json":"https://pith.science/api/pith-number/XZ7GRVPQIBJOJPG3DLGAU6RYJB/graph.json","events_json":"https://pith.science/api/pith-number/XZ7GRVPQIBJOJPG3DLGAU6RYJB/events.json","paper":"https://pith.science/paper/XZ7GRVPQ"},"agent_actions":{"view_html":"https://pith.science/pith/XZ7GRVPQIBJOJPG3DLGAU6RYJB","download_json":"https://pith.science/pith/XZ7GRVPQIBJOJPG3DLGAU6RYJB.json","view_paper":"https://pith.science/paper/XZ7GRVPQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.11646&json=true","fetch_graph":"https://pith.science/api/pith-number/XZ7GRVPQIBJOJPG3DLGAU6RYJB/graph.json","fetch_events":"https://pith.science/api/pith-number/XZ7GRVPQIBJOJPG3DLGAU6RYJB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XZ7GRVPQIBJOJPG3DLGAU6RYJB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XZ7GRVPQIBJOJPG3DLGAU6RYJB/action/storage_attestation","attest_author":"https://pith.science/pith/XZ7GRVPQIBJOJPG3DLGAU6RYJB/action/author_attestation","sign_citation":"https://pith.science/pith/XZ7GRVPQIBJOJPG3DLGAU6RYJB/action/citation_signature","submit_replication":"https://pith.science/pith/XZ7GRVPQIBJOJPG3DLGAU6RYJB/action/replication_record"}},"created_at":"2026-07-05T10:31:36.609494+00:00","updated_at":"2026-07-05T10:31:36.609494+00:00"}