{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:2RLM54QUMQIAGSGV4QREL64GIK","short_pith_number":"pith:2RLM54QU","schema_version":"1.0","canonical_sha256":"d456cef21464100348d5e42245fb8642a69dcd168bc9ea43afc7d637366b7aef","source":{"kind":"arxiv","id":"2605.15523","version":1},"attestation_state":"computed","paper":{"title":"Self-Prompting Diffusion Transformer for Open-Vocabulary Scene Text Editing via In-Context Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Self-prompting constructs style and glyph prompts directly from the source image so a Multi-Modal Diffusion Transformer can edit scene text in any vocabulary while preserving original appearance.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chengjing Wu, Hongxi Li, Jiangtao Yao, Luoqi Liu, Tianbao Liu, Ting Liu, Tong Wang, Xiaochao Qu, Xinxiao Wu","submitted_at":"2026-05-15T01:44:17Z","abstract_excerpt":"Scene text editing aims to modify text in a target region of an image while preserving surrounding background style and texture. Existing methods rely solely on image background information while neglecting the visual details of target regions, which discards stylistic features in the original text and essentially degrades the task to text rendering. Moreover, the conditions imposed by pre-trained glyph encoder limit the scope of editable text. To address these issues, this paper proposes a self-prompting scene text editing method that constructs style and glyph prompts directly from the origi"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2605.15523","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-15T01:44:17Z","cross_cats_sorted":[],"title_canon_sha256":"fdae8ebe27f016fd7ed616ab4d3679e9592394b98a3bd4d2a7098519cb8d025e","abstract_canon_sha256":"233ff028ff8ed25410bed90d708c1883da74eb36a768888cd71e79ab295e8355"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:01:03.144962Z","signature_b64":"r3LxovKtMWFKwwFWugOKubGScH8GTFShNb7APzjH91Dl0dMdlFK0xtvq5QIKinyyF8FSHKIf1dqvpXhlqfFBBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d456cef21464100348d5e42245fb8642a69dcd168bc9ea43afc7d637366b7aef","last_reissued_at":"2026-05-20T00:01:03.144075Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:01:03.144075Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Self-Prompting Diffusion Transformer for Open-Vocabulary Scene Text Editing via In-Context Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Self-prompting constructs style and glyph prompts directly from the source image so a Multi-Modal Diffusion Transformer can edit scene text in any vocabulary while preserving original appearance.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chengjing Wu, Hongxi Li, Jiangtao Yao, Luoqi Liu, Tianbao Liu, Ting Liu, Tong Wang, Xiaochao Qu, Xinxiao Wu","submitted_at":"2026-05-15T01:44:17Z","abstract_excerpt":"Scene text editing aims to modify text in a target region of an image while preserving surrounding background style and texture. Existing methods rely solely on image background information while neglecting the visual details of target regions, which discards stylistic features in the original text and essentially degrades the task to text rendering. Moreover, the conditions imposed by pre-trained glyph encoder limit the scope of editable text. To address these issues, this paper proposes a self-prompting scene text editing method that constructs style and glyph prompts directly from the origi"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"By leveraging the in-context learning capability of the Multi-Modal Diffusion Transformer (MM-DiT), it achieves open-vocabulary and style-consistent text editing. Experimental results on various languages demonstrate that our method achieves the state-of-the-art performance in both text accuracy and style consistency.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That prompts constructed directly from the original image via self-prompting are sufficient to capture all stylistic and glyph details without any additional dedicated encoders or external conditioning.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A self-prompting MM-DiT model performs open-vocabulary scene text editing by extracting style and glyph information from the original image without extra encoders.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Self-prompting constructs style and glyph prompts directly from the source image so a Multi-Modal Diffusion Transformer can edit scene text in any vocabulary while preserving original appearance.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"5658923bf21fe5d1f647802b6eedc2adca61be1074c843501a8593487009daee"},"source":{"id":"2605.15523","kind":"arxiv","version":1},"verdict":{"id":"a40d1d16-8940-42b3-ab95-e04c03f6dab0","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T15:01:51.225696Z","strongest_claim":"By leveraging the in-context learning capability of the Multi-Modal Diffusion Transformer (MM-DiT), it achieves open-vocabulary and style-consistent text editing. Experimental results on various languages demonstrate that our method achieves the state-of-the-art performance in both text accuracy and style consistency.","one_line_summary":"A self-prompting MM-DiT model performs open-vocabulary scene text editing by extracting style and glyph information from the original image without extra encoders.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That prompts constructed directly from the original image via self-prompting are sufficient to capture all stylistic and glyph details without any additional dedicated encoders or external conditioning.","pith_extraction_headline":"Self-prompting constructs style and glyph prompts directly from the source image so a Multi-Modal Diffusion Transformer can edit scene text in any vocabulary while preserving original appearance."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.15523/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T15:31:17.683711Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T15:10:43.396323Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"cited_work_retraction","ran_at":"2026-05-19T14:22:02.935565Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T14:21:54.044469Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"shingle_duplication","ran_at":"2026-05-19T13:49:41.841182Z","status":"skipped","version":"0.1.0","findings_count":0},{"name":"citation_quote_validity","ran_at":"2026-05-19T13:49:41.378455Z","status":"skipped","version":"0.1.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T13:33:22.626626Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"e547add522e636866b3a6a007f59326029c2aaf3f1d0aaea1a5ed23cbbbbff4a"},"references":{"count":13,"sample":[{"doi":"","year":2009,"title":"PP-OCR: A practical ultra lightweight OCR system.CoRR, abs/2009.09941","work_id":"555e4547-6e8f-49f5-acf4-0adae9b359e3","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Metadata conditioning accelerates language model pre-training.arXiv preprint arXiv:2501.01956,","work_id":"1c546854-ed6f-4b2a-98b2-0e9b024d2c89","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Flux-text: A simple and advanced diffusion transformer baseline for scene text editing","work_id":"1bf6b24b-cdb9-4b2d-b872-9613d9085131","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Rustitw: Russian language text dataset for visual text in- the-wild recognition.arXiv preprint arXiv:2303.16531,","work_id":"18510534-e21b-4d5a-a6f7-02a455c17cdb","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2019,"title":"N., Karatzas, D., Khlif, W., Matas, J., Pal, U., Burie, J.-C., Liu, C.-l., et al","work_id":"bacd82f5-b12f-495e-94a0-bf1368d6fea1","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":13,"snapshot_sha256":"a5cd98eaa5a7fbe4d6e49d09996ab80efed6ad2a7cc66c4768d05415f6dde81d","internal_anchors":2},"formal_canon":{"evidence_count":2,"snapshot_sha256":"3aecd0f4c9eb93e20b2503f734dc39a17f030ba7bd6cd46f6032a2ea2ad7ffd2"},"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":"2605.15523","created_at":"2026-05-20T00:01:03.144224+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.15523v1","created_at":"2026-05-20T00:01:03.144224+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.15523","created_at":"2026-05-20T00:01:03.144224+00:00"},{"alias_kind":"pith_short_12","alias_value":"2RLM54QUMQIA","created_at":"2026-05-20T00:01:03.144224+00:00"},{"alias_kind":"pith_short_16","alias_value":"2RLM54QUMQIAGSGV","created_at":"2026-05-20T00:01:03.144224+00:00"},{"alias_kind":"pith_short_8","alias_value":"2RLM54QU","created_at":"2026-05-20T00:01:03.144224+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2RLM54QUMQIAGSGV4QREL64GIK","json":"https://pith.science/pith/2RLM54QUMQIAGSGV4QREL64GIK.json","graph_json":"https://pith.science/api/pith-number/2RLM54QUMQIAGSGV4QREL64GIK/graph.json","events_json":"https://pith.science/api/pith-number/2RLM54QUMQIAGSGV4QREL64GIK/events.json","paper":"https://pith.science/paper/2RLM54QU"},"agent_actions":{"view_html":"https://pith.science/pith/2RLM54QUMQIAGSGV4QREL64GIK","download_json":"https://pith.science/pith/2RLM54QUMQIAGSGV4QREL64GIK.json","view_paper":"https://pith.science/paper/2RLM54QU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.15523&json=true","fetch_graph":"https://pith.science/api/pith-number/2RLM54QUMQIAGSGV4QREL64GIK/graph.json","fetch_events":"https://pith.science/api/pith-number/2RLM54QUMQIAGSGV4QREL64GIK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2RLM54QUMQIAGSGV4QREL64GIK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2RLM54QUMQIAGSGV4QREL64GIK/action/storage_attestation","attest_author":"https://pith.science/pith/2RLM54QUMQIAGSGV4QREL64GIK/action/author_attestation","sign_citation":"https://pith.science/pith/2RLM54QUMQIAGSGV4QREL64GIK/action/citation_signature","submit_replication":"https://pith.science/pith/2RLM54QUMQIAGSGV4QREL64GIK/action/replication_record"}},"created_at":"2026-05-20T00:01:03.144224+00:00","updated_at":"2026-05-20T00:01:03.144224+00:00"}