{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:GIMYHKSJZF6OHLX2YWHWA5XXGG","short_pith_number":"pith:GIMYHKSJ","schema_version":"1.0","canonical_sha256":"321983aa49c97ce3aefac58f6076f731af4e39ec9f3b166ac294592be82b3cff","source":{"kind":"arxiv","id":"2605.28456","version":1},"attestation_state":"computed","paper":{"title":"Diffusion Large Language Models for Visual Speech Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","eess.AS"],"primary_cat":"cs.AI","authors_text":"Chae Won Kim, Hyeongseop Rha, Jeong Hun Yeo, Yong Man Ro","submitted_at":"2026-05-27T13:22:08Z","abstract_excerpt":"Existing Visual Speech Recognition (VSR) systems commonly rely on left-to-right autoregressive decoding, which can force premature decisions on visually ambiguous tokens before sufficient context is available. We propose DLLM-VSR, to the best of our knowledge, the first Diffusion Large Language Model (DLLM)-based VSR framework, formulating transcription as iterative masked denoising with flexible-order decoding. With confidence-based unmasking, DLLM-VSR commits high-confidence positions early and uses the committed tokens as bidirectional context to refine ambiguous ones. To adapt DLLMs to VSR"},"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":"2605.28456","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-27T13:22:08Z","cross_cats_sorted":["cs.CV","eess.AS"],"title_canon_sha256":"5f1800b07298735af9d0b29bb58999b7b6641ba4de3dc663eca4893f3adb2f45","abstract_canon_sha256":"ea47e2dc3e6b8703ab03bf11d3ca4bf269674b5b503e7f463345d3d230eac4f5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-28T02:04:53.551155Z","signature_b64":"vssPXqWxo6XVN1pFUj1zKVGUCSP6GehgIpOKQNjpLEP6j0sW79Yd1wMNzqsoIFnZIUZhyNjzVowaVfYu1qj0Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"321983aa49c97ce3aefac58f6076f731af4e39ec9f3b166ac294592be82b3cff","last_reissued_at":"2026-05-28T02:04:53.550659Z","signature_status":"signed_v1","first_computed_at":"2026-05-28T02:04:53.550659Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Diffusion Large Language Models for Visual Speech Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","eess.AS"],"primary_cat":"cs.AI","authors_text":"Chae Won Kim, Hyeongseop Rha, Jeong Hun Yeo, Yong Man Ro","submitted_at":"2026-05-27T13:22:08Z","abstract_excerpt":"Existing Visual Speech Recognition (VSR) systems commonly rely on left-to-right autoregressive decoding, which can force premature decisions on visually ambiguous tokens before sufficient context is available. We propose DLLM-VSR, to the best of our knowledge, the first Diffusion Large Language Model (DLLM)-based VSR framework, formulating transcription as iterative masked denoising with flexible-order decoding. With confidence-based unmasking, DLLM-VSR commits high-confidence positions early and uses the committed tokens as bidirectional context to refine ambiguous ones. To adapt DLLMs to VSR"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.28456","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/2605.28456/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":"2605.28456","created_at":"2026-05-28T02:04:53.550729+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.28456v1","created_at":"2026-05-28T02:04:53.550729+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.28456","created_at":"2026-05-28T02:04:53.550729+00:00"},{"alias_kind":"pith_short_12","alias_value":"GIMYHKSJZF6O","created_at":"2026-05-28T02:04:53.550729+00:00"},{"alias_kind":"pith_short_16","alias_value":"GIMYHKSJZF6OHLX2","created_at":"2026-05-28T02:04:53.550729+00:00"},{"alias_kind":"pith_short_8","alias_value":"GIMYHKSJ","created_at":"2026-05-28T02:04:53.550729+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GIMYHKSJZF6OHLX2YWHWA5XXGG","json":"https://pith.science/pith/GIMYHKSJZF6OHLX2YWHWA5XXGG.json","graph_json":"https://pith.science/api/pith-number/GIMYHKSJZF6OHLX2YWHWA5XXGG/graph.json","events_json":"https://pith.science/api/pith-number/GIMYHKSJZF6OHLX2YWHWA5XXGG/events.json","paper":"https://pith.science/paper/GIMYHKSJ"},"agent_actions":{"view_html":"https://pith.science/pith/GIMYHKSJZF6OHLX2YWHWA5XXGG","download_json":"https://pith.science/pith/GIMYHKSJZF6OHLX2YWHWA5XXGG.json","view_paper":"https://pith.science/paper/GIMYHKSJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.28456&json=true","fetch_graph":"https://pith.science/api/pith-number/GIMYHKSJZF6OHLX2YWHWA5XXGG/graph.json","fetch_events":"https://pith.science/api/pith-number/GIMYHKSJZF6OHLX2YWHWA5XXGG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GIMYHKSJZF6OHLX2YWHWA5XXGG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GIMYHKSJZF6OHLX2YWHWA5XXGG/action/storage_attestation","attest_author":"https://pith.science/pith/GIMYHKSJZF6OHLX2YWHWA5XXGG/action/author_attestation","sign_citation":"https://pith.science/pith/GIMYHKSJZF6OHLX2YWHWA5XXGG/action/citation_signature","submit_replication":"https://pith.science/pith/GIMYHKSJZF6OHLX2YWHWA5XXGG/action/replication_record"}},"created_at":"2026-05-28T02:04:53.550729+00:00","updated_at":"2026-05-28T02:04:53.550729+00:00"}