{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:DV7PNE2RDHZDMGNUBDBVWK4XAP","short_pith_number":"pith:DV7PNE2R","schema_version":"1.0","canonical_sha256":"1d7ef6935119f23619b408c35b2b9703df5d3df347cace3de52490fa3ca14092","source":{"kind":"arxiv","id":"2505.02823","version":1},"attestation_state":"computed","paper":{"title":"MUSAR: Exploring Multi-Subject Customization from Single-Subject Dataset via Attention Routing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chong Mou, Pengze Zhang, Qian He, Songtao Zhao, Yanze Wu, Zinan Guo","submitted_at":"2025-05-05T17:50:24Z","abstract_excerpt":"Current multi-subject customization approaches encounter two critical challenges: the difficulty in acquiring diverse multi-subject training data, and attribute entanglement across different subjects. To bridge these gaps, we propose MUSAR - a simple yet effective framework to achieve robust multi-subject customization while requiring only single-subject training data. Firstly, to break the data limitation, we introduce debiased diptych learning. It constructs diptych training pairs from single-subject images to facilitate multi-subject learning, while actively correcting the distribution bias"},"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":"2505.02823","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-05-05T17:50:24Z","cross_cats_sorted":[],"title_canon_sha256":"4c486620ad2631839cad81a1cf3967132f43fb6a14878c7d26720bbb877db0e7","abstract_canon_sha256":"7b31be8ba0173bbb0e3357e5ff826eb8c8d54b4b7c54ac632f4b6766ccc57363"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:58:43.384155Z","signature_b64":"Qh5Z7xS5N4rbi2DMGiGN+ZdJmvp45UUOBMkljPzTx5Q5sl+5g5G6jff1tupi+OlEVOF6SwQ3Fscxz/2dsX9KAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1d7ef6935119f23619b408c35b2b9703df5d3df347cace3de52490fa3ca14092","last_reissued_at":"2026-07-05T10:58:43.383643Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:58:43.383643Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MUSAR: Exploring Multi-Subject Customization from Single-Subject Dataset via Attention Routing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chong Mou, Pengze Zhang, Qian He, Songtao Zhao, Yanze Wu, Zinan Guo","submitted_at":"2025-05-05T17:50:24Z","abstract_excerpt":"Current multi-subject customization approaches encounter two critical challenges: the difficulty in acquiring diverse multi-subject training data, and attribute entanglement across different subjects. To bridge these gaps, we propose MUSAR - a simple yet effective framework to achieve robust multi-subject customization while requiring only single-subject training data. Firstly, to break the data limitation, we introduce debiased diptych learning. It constructs diptych training pairs from single-subject images to facilitate multi-subject learning, while actively correcting the distribution bias"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.02823","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/2505.02823/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":"2505.02823","created_at":"2026-07-05T10:58:43.383702+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.02823v1","created_at":"2026-07-05T10:58:43.383702+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.02823","created_at":"2026-07-05T10:58:43.383702+00:00"},{"alias_kind":"pith_short_12","alias_value":"DV7PNE2RDHZD","created_at":"2026-07-05T10:58:43.383702+00:00"},{"alias_kind":"pith_short_16","alias_value":"DV7PNE2RDHZDMGNU","created_at":"2026-07-05T10:58:43.383702+00:00"},{"alias_kind":"pith_short_8","alias_value":"DV7PNE2R","created_at":"2026-07-05T10:58:43.383702+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26058","citing_title":"DomainShuttle: Freeform Open Domain Subject-driven Text-to-video Generation","ref_index":51,"is_internal_anchor":false},{"citing_arxiv_id":"2512.12675","citing_title":"Scone: Bridging Composition and Distinction in Subject-Driven Image Generation via Unified Understanding-Generation Modeling","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DV7PNE2RDHZDMGNUBDBVWK4XAP","json":"https://pith.science/pith/DV7PNE2RDHZDMGNUBDBVWK4XAP.json","graph_json":"https://pith.science/api/pith-number/DV7PNE2RDHZDMGNUBDBVWK4XAP/graph.json","events_json":"https://pith.science/api/pith-number/DV7PNE2RDHZDMGNUBDBVWK4XAP/events.json","paper":"https://pith.science/paper/DV7PNE2R"},"agent_actions":{"view_html":"https://pith.science/pith/DV7PNE2RDHZDMGNUBDBVWK4XAP","download_json":"https://pith.science/pith/DV7PNE2RDHZDMGNUBDBVWK4XAP.json","view_paper":"https://pith.science/paper/DV7PNE2R","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.02823&json=true","fetch_graph":"https://pith.science/api/pith-number/DV7PNE2RDHZDMGNUBDBVWK4XAP/graph.json","fetch_events":"https://pith.science/api/pith-number/DV7PNE2RDHZDMGNUBDBVWK4XAP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DV7PNE2RDHZDMGNUBDBVWK4XAP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DV7PNE2RDHZDMGNUBDBVWK4XAP/action/storage_attestation","attest_author":"https://pith.science/pith/DV7PNE2RDHZDMGNUBDBVWK4XAP/action/author_attestation","sign_citation":"https://pith.science/pith/DV7PNE2RDHZDMGNUBDBVWK4XAP/action/citation_signature","submit_replication":"https://pith.science/pith/DV7PNE2RDHZDMGNUBDBVWK4XAP/action/replication_record"}},"created_at":"2026-07-05T10:58:43.383702+00:00","updated_at":"2026-07-05T10:58:43.383702+00:00"}