{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:PW242HYTT634F5JTLHF6URLWNQ","short_pith_number":"pith:PW242HYT","schema_version":"1.0","canonical_sha256":"7db5cd1f139fb7c2f53359cbea45766c0bd86e6c3895ea530480ac95d1fa50f1","source":{"kind":"arxiv","id":"2608.15690","version":1},"attestation_state":"computed","paper":{"title":"Adding Voice Cloning to Text-to-Audio-Video Models with a Single Zero-Initialised Layer","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.MM"],"primary_cat":"cs.SD","authors_text":"Alexey Letunovskiy, Anna Dmitrienko, Denis Dimitrov, Ivan Kirillov, Ivan Mikheev, Kirill Chernyshev, Viacheslav Vasilev","submitted_at":"2026-08-16T11:40:20Z","abstract_excerpt":"Text-to-audio-video (T2AV) generation models produce a video and its soundtrack from a textual description, but offer no control over whose voice speaks in the output. We show that a base T2AV model can be turned into a voice-cloning model by adding a single zero-initialized linear layer on top of its audio backbone, fine-tuning for a comparatively short training schedule, and conditioning on a short reference recording at inference time. The reference is injected through two complementary signals: its diffusion latents are prepended to the audio stream, and a global speaker embedding modulate"},"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":"2608.15690","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SD","submitted_at":"2026-08-16T11:40:20Z","cross_cats_sorted":["cs.AI","cs.LG","cs.MM"],"title_canon_sha256":"66055f8e9bea9f21a0dee839f43c6313567c13f6a2a87f23a345f969bc8d6adf","abstract_canon_sha256":"d7f67596d7a0a5de1a1f22536c98cea2b750ec4c5d0a703639f83ce1cf6ea6c4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-18T01:15:59.302271Z","signature_b64":"own0Qpp9MJ2+tGUR/ncfy1JtL4Et8FNx/SpBk1vVok/HYrId/IBVp2x82OQMCmauYsCPt2yrYVMa9iDXaI51AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7db5cd1f139fb7c2f53359cbea45766c0bd86e6c3895ea530480ac95d1fa50f1","last_reissued_at":"2026-08-18T01:15:59.300789Z","signature_status":"signed_v1","first_computed_at":"2026-08-18T01:15:59.300789Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Adding Voice Cloning to Text-to-Audio-Video Models with a Single Zero-Initialised Layer","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.MM"],"primary_cat":"cs.SD","authors_text":"Alexey Letunovskiy, Anna Dmitrienko, Denis Dimitrov, Ivan Kirillov, Ivan Mikheev, Kirill Chernyshev, Viacheslav Vasilev","submitted_at":"2026-08-16T11:40:20Z","abstract_excerpt":"Text-to-audio-video (T2AV) generation models produce a video and its soundtrack from a textual description, but offer no control over whose voice speaks in the output. We show that a base T2AV model can be turned into a voice-cloning model by adding a single zero-initialized linear layer on top of its audio backbone, fine-tuning for a comparatively short training schedule, and conditioning on a short reference recording at inference time. The reference is injected through two complementary signals: its diffusion latents are prepended to the audio stream, and a global speaker embedding modulate"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.15690","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/2608.15690/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":"2608.15690","created_at":"2026-08-18T01:15:59.301308+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.15690v1","created_at":"2026-08-18T01:15:59.301308+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.15690","created_at":"2026-08-18T01:15:59.301308+00:00"},{"alias_kind":"pith_short_12","alias_value":"PW242HYTT634","created_at":"2026-08-18T01:15:59.301308+00:00"},{"alias_kind":"pith_short_16","alias_value":"PW242HYTT634F5JT","created_at":"2026-08-18T01:15:59.301308+00:00"},{"alias_kind":"pith_short_8","alias_value":"PW242HYT","created_at":"2026-08-18T01:15:59.301308+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/PW242HYTT634F5JTLHF6URLWNQ","json":"https://pith.science/pith/PW242HYTT634F5JTLHF6URLWNQ.json","graph_json":"https://pith.science/api/pith-number/PW242HYTT634F5JTLHF6URLWNQ/graph.json","events_json":"https://pith.science/api/pith-number/PW242HYTT634F5JTLHF6URLWNQ/events.json","paper":"https://pith.science/paper/PW242HYT"},"agent_actions":{"view_html":"https://pith.science/pith/PW242HYTT634F5JTLHF6URLWNQ","download_json":"https://pith.science/pith/PW242HYTT634F5JTLHF6URLWNQ.json","view_paper":"https://pith.science/paper/PW242HYT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.15690&json=true","fetch_graph":"https://pith.science/api/pith-number/PW242HYTT634F5JTLHF6URLWNQ/graph.json","fetch_events":"https://pith.science/api/pith-number/PW242HYTT634F5JTLHF6URLWNQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PW242HYTT634F5JTLHF6URLWNQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PW242HYTT634F5JTLHF6URLWNQ/action/storage_attestation","attest_author":"https://pith.science/pith/PW242HYTT634F5JTLHF6URLWNQ/action/author_attestation","sign_citation":"https://pith.science/pith/PW242HYTT634F5JTLHF6URLWNQ/action/citation_signature","submit_replication":"https://pith.science/pith/PW242HYTT634F5JTLHF6URLWNQ/action/replication_record"}},"created_at":"2026-08-18T01:15:59.301308+00:00","updated_at":"2026-08-18T01:15:59.301308+00:00"}