{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XED5PYL4PN35RCO363OIG6D5HM","short_pith_number":"pith:XED5PYL4","schema_version":"1.0","canonical_sha256":"b907d7e17c7b77d889dbf6dc83787d3b0f6e2ffa42e74ba7d973168e4b933bc3","source":{"kind":"arxiv","id":"2504.00396","version":3},"attestation_state":"computed","paper":{"title":"SPF-Portrait: Towards Pure Text-to-Portrait Customization with Semantic Pollution-Free Fine-Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Linlin Shen, Long Zeng, Pengfei Wan, Pingfa Feng, Qingyu Li, Weicheng Xie, Wenyu Qin, Xiaole Xian, Zhichao Liao","submitted_at":"2025-04-01T03:37:30Z","abstract_excerpt":"Fine-tuning a pre-trained Text-to-Image (T2I) model on a tailored portrait dataset is the mainstream method for text-to-portrait customization. However, existing methods often severely impact the original model's behavior (e.g., changes in ID, layout, etc.) while customizing portrait attributes. To address this issue, we propose SPF-Portrait, a pioneering work to purely understand customized target semantics and minimize disruption to the original model. In our SPF-Portrait, we design a dual-path contrastive learning pipeline, which introduces the original model as a behavioral alignment refer"},"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":"2504.00396","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-04-01T03:37:30Z","cross_cats_sorted":[],"title_canon_sha256":"aad6994490e10bdc8b71571edc773d0f7a6397ecade9ce2fda4f02842aff8dcb","abstract_canon_sha256":"5128360be4dd96921b0750d0e7e2cd2396f3614dd6d179921f680a4a8e267c21"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:10:36.688382Z","signature_b64":"Ks1b3E5W/ZDOI2CogS1qavG2cd/bA9BZ79es6jK1SMMqwhpwShWGe+SO2qb3DVqTg92mhb+IBs40BU2xFv2QAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b907d7e17c7b77d889dbf6dc83787d3b0f6e2ffa42e74ba7d973168e4b933bc3","last_reissued_at":"2026-07-05T11:10:36.687874Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:10:36.687874Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SPF-Portrait: Towards Pure Text-to-Portrait Customization with Semantic Pollution-Free Fine-Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Linlin Shen, Long Zeng, Pengfei Wan, Pingfa Feng, Qingyu Li, Weicheng Xie, Wenyu Qin, Xiaole Xian, Zhichao Liao","submitted_at":"2025-04-01T03:37:30Z","abstract_excerpt":"Fine-tuning a pre-trained Text-to-Image (T2I) model on a tailored portrait dataset is the mainstream method for text-to-portrait customization. However, existing methods often severely impact the original model's behavior (e.g., changes in ID, layout, etc.) while customizing portrait attributes. To address this issue, we propose SPF-Portrait, a pioneering work to purely understand customized target semantics and minimize disruption to the original model. In our SPF-Portrait, we design a dual-path contrastive learning pipeline, which introduces the original model as a behavioral alignment refer"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.00396","kind":"arxiv","version":3},"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/2504.00396/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":"2504.00396","created_at":"2026-07-05T11:10:36.687928+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.00396v3","created_at":"2026-07-05T11:10:36.687928+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.00396","created_at":"2026-07-05T11:10:36.687928+00:00"},{"alias_kind":"pith_short_12","alias_value":"XED5PYL4PN35","created_at":"2026-07-05T11:10:36.687928+00:00"},{"alias_kind":"pith_short_16","alias_value":"XED5PYL4PN35RCO3","created_at":"2026-07-05T11:10:36.687928+00:00"},{"alias_kind":"pith_short_8","alias_value":"XED5PYL4","created_at":"2026-07-05T11:10:36.687928+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2603.07561","citing_title":"PureCC: Pure Learning for Text-to-Image Concept Customization","ref_index":48,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XED5PYL4PN35RCO363OIG6D5HM","json":"https://pith.science/pith/XED5PYL4PN35RCO363OIG6D5HM.json","graph_json":"https://pith.science/api/pith-number/XED5PYL4PN35RCO363OIG6D5HM/graph.json","events_json":"https://pith.science/api/pith-number/XED5PYL4PN35RCO363OIG6D5HM/events.json","paper":"https://pith.science/paper/XED5PYL4"},"agent_actions":{"view_html":"https://pith.science/pith/XED5PYL4PN35RCO363OIG6D5HM","download_json":"https://pith.science/pith/XED5PYL4PN35RCO363OIG6D5HM.json","view_paper":"https://pith.science/paper/XED5PYL4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.00396&json=true","fetch_graph":"https://pith.science/api/pith-number/XED5PYL4PN35RCO363OIG6D5HM/graph.json","fetch_events":"https://pith.science/api/pith-number/XED5PYL4PN35RCO363OIG6D5HM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XED5PYL4PN35RCO363OIG6D5HM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XED5PYL4PN35RCO363OIG6D5HM/action/storage_attestation","attest_author":"https://pith.science/pith/XED5PYL4PN35RCO363OIG6D5HM/action/author_attestation","sign_citation":"https://pith.science/pith/XED5PYL4PN35RCO363OIG6D5HM/action/citation_signature","submit_replication":"https://pith.science/pith/XED5PYL4PN35RCO363OIG6D5HM/action/replication_record"}},"created_at":"2026-07-05T11:10:36.687928+00:00","updated_at":"2026-07-05T11:10:36.687928+00:00"}