{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:SXGMN2CRVSJAIEQ73EGTOOH64S","short_pith_number":"pith:SXGMN2CR","schema_version":"1.0","canonical_sha256":"95ccc6e851ac9204121fd90d3738fee4a409cb63a610e402f772ab959750e6f5","source":{"kind":"arxiv","id":"2403.16973","version":3},"attestation_state":"computed","paper":{"title":"VoiceCraft: Zero-Shot Speech Editing and Text-to-Speech in the Wild","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Abdelrahman Mohamed, David Harwath, Po-Yao Huang, Puyuan Peng, Shang-Wen Li","submitted_at":"2024-03-25T17:38:32Z","abstract_excerpt":"We introduce VoiceCraft, a token infilling neural codec language model, that achieves state-of-the-art performance on both speech editing and zero-shot text-to-speech (TTS) on audiobooks, internet videos, and podcasts. VoiceCraft employs a Transformer decoder architecture and introduces a token rearrangement procedure that combines causal masking and delayed stacking to enable generation within an existing sequence. On speech editing tasks, VoiceCraft produces edited speech that is nearly indistinguishable from unedited recordings in terms of naturalness, as evaluated by humans; for zero-shot "},"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":"2403.16973","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"eess.AS","submitted_at":"2024-03-25T17:38:32Z","cross_cats_sorted":["cs.AI","cs.CL","cs.LG","cs.SD"],"title_canon_sha256":"b524d510e3916dfaacc9498003ec01ef09633c30b9de0f56d5412416823d8182","abstract_canon_sha256":"872ce33d97ce600bc34e9685667491d5caa3a678ca9b5f230ef1bece6736c5de"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:31:49.616480Z","signature_b64":"+9J3ih7eu/XKSo03Q1EXx3SL6PjlTAR6gpxFMewbzi3u2QkIB/+ZXYQ0n9FQDvCaCWUnMEcxoFebcePhq963AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"95ccc6e851ac9204121fd90d3738fee4a409cb63a610e402f772ab959750e6f5","last_reissued_at":"2026-07-05T08:31:49.615968Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:31:49.615968Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"VoiceCraft: Zero-Shot Speech Editing and Text-to-Speech in the Wild","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Abdelrahman Mohamed, David Harwath, Po-Yao Huang, Puyuan Peng, Shang-Wen Li","submitted_at":"2024-03-25T17:38:32Z","abstract_excerpt":"We introduce VoiceCraft, a token infilling neural codec language model, that achieves state-of-the-art performance on both speech editing and zero-shot text-to-speech (TTS) on audiobooks, internet videos, and podcasts. VoiceCraft employs a Transformer decoder architecture and introduces a token rearrangement procedure that combines causal masking and delayed stacking to enable generation within an existing sequence. On speech editing tasks, VoiceCraft produces edited speech that is nearly indistinguishable from unedited recordings in terms of naturalness, as evaluated by humans; for zero-shot "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.16973","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/2403.16973/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":"2403.16973","created_at":"2026-07-05T08:31:49.616036+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.16973v3","created_at":"2026-07-05T08:31:49.616036+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.16973","created_at":"2026-07-05T08:31:49.616036+00:00"},{"alias_kind":"pith_short_12","alias_value":"SXGMN2CRVSJA","created_at":"2026-07-05T08:31:49.616036+00:00"},{"alias_kind":"pith_short_16","alias_value":"SXGMN2CRVSJAIEQ7","created_at":"2026-07-05T08:31:49.616036+00:00"},{"alias_kind":"pith_short_8","alias_value":"SXGMN2CR","created_at":"2026-07-05T08:31:49.616036+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.31247","citing_title":"FlexiSLM: A Dynamic and Controllable Frame Rate Spoken Language Model","ref_index":127,"is_internal_anchor":false},{"citing_arxiv_id":"2506.23552","citing_title":"JAM-Flow: Joint Audio-Motion Synthesis with Flow Matching","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2410.06885","citing_title":"F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching","ref_index":130,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11917","citing_title":"StreamMark: A Deep Learning-Based Semi-Fragile Audio Watermarking for Proactive Deepfake Detection","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11283","citing_title":"Multimodal Large Language Model-Enabled Video Translation: A Role-Oriented Survey","ref_index":108,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SXGMN2CRVSJAIEQ73EGTOOH64S","json":"https://pith.science/pith/SXGMN2CRVSJAIEQ73EGTOOH64S.json","graph_json":"https://pith.science/api/pith-number/SXGMN2CRVSJAIEQ73EGTOOH64S/graph.json","events_json":"https://pith.science/api/pith-number/SXGMN2CRVSJAIEQ73EGTOOH64S/events.json","paper":"https://pith.science/paper/SXGMN2CR"},"agent_actions":{"view_html":"https://pith.science/pith/SXGMN2CRVSJAIEQ73EGTOOH64S","download_json":"https://pith.science/pith/SXGMN2CRVSJAIEQ73EGTOOH64S.json","view_paper":"https://pith.science/paper/SXGMN2CR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.16973&json=true","fetch_graph":"https://pith.science/api/pith-number/SXGMN2CRVSJAIEQ73EGTOOH64S/graph.json","fetch_events":"https://pith.science/api/pith-number/SXGMN2CRVSJAIEQ73EGTOOH64S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SXGMN2CRVSJAIEQ73EGTOOH64S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SXGMN2CRVSJAIEQ73EGTOOH64S/action/storage_attestation","attest_author":"https://pith.science/pith/SXGMN2CRVSJAIEQ73EGTOOH64S/action/author_attestation","sign_citation":"https://pith.science/pith/SXGMN2CRVSJAIEQ73EGTOOH64S/action/citation_signature","submit_replication":"https://pith.science/pith/SXGMN2CRVSJAIEQ73EGTOOH64S/action/replication_record"}},"created_at":"2026-07-05T08:31:49.616036+00:00","updated_at":"2026-07-05T08:31:49.616036+00:00"}