{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4RH6VKKNWUKIS7F6O5I52DE77O","short_pith_number":"pith:4RH6VKKN","schema_version":"1.0","canonical_sha256":"e44feaa94db514897cbe7751dd0c9ffbb8ffaca3de1c05a4bf3823ef2c63e0b2","source":{"kind":"arxiv","id":"2606.29031","version":1},"attestation_state":"computed","paper":{"title":"How to Leverage Synthetic Speech for LLM-Based ASR Systems?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Andreas Stolcke, Dairazalia Sanchez-Cortes, Esa\\'u Villatoro-Tello, Kadri Hacio\\u{g}lu, Manjunath K E, Old\\v{r}ich Plchot, Petr Motlicek, Sergio Burdisso, S\\'everin Baroudi, Shashi Kumar, Srikanth Madikeri, Yanis Labrak","submitted_at":"2026-06-27T17:57:27Z","abstract_excerpt":"In regulated domains such as banking and healthcare, where privacy constraints make real speech costly to collect and retain, synthetic speech from modern text-to-speech (TTS) is an appealing alternative for training automatic speech recognition (ASR) without exposing sensitive customer recordings. Yet a persistent distributional gap between synthetic and real data limits how far it can replace genuine recordings. Prior work largely treats this gap as a black box to be engineered around, but in our work, we instead examine its origin directly by probing a SLAM-ASR architecture. Then, we locali"},"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":"2606.29031","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-27T17:57:27Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"55431eb1a0aa2e1e98c379a1b9ec1617a4180b7ac71c908342aefca067b00c59","abstract_canon_sha256":"3a6031580bdcb4527ec035377ba6a5d200473eefd98cc5e1e0f6be3f9c47469b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-30T01:17:50.091969Z","signature_b64":"0Tb0wGBCQpZr4khzAQgOpQvW7sAMPn2etde32C2psoU2glBBtCdu86C88ZKzae+yQGFTwo5w8iosWwkBBjzpCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e44feaa94db514897cbe7751dd0c9ffbb8ffaca3de1c05a4bf3823ef2c63e0b2","last_reissued_at":"2026-06-30T01:17:50.091294Z","signature_status":"signed_v1","first_computed_at":"2026-06-30T01:17:50.091294Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How to Leverage Synthetic Speech for LLM-Based ASR Systems?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Andreas Stolcke, Dairazalia Sanchez-Cortes, Esa\\'u Villatoro-Tello, Kadri Hacio\\u{g}lu, Manjunath K E, Old\\v{r}ich Plchot, Petr Motlicek, Sergio Burdisso, S\\'everin Baroudi, Shashi Kumar, Srikanth Madikeri, Yanis Labrak","submitted_at":"2026-06-27T17:57:27Z","abstract_excerpt":"In regulated domains such as banking and healthcare, where privacy constraints make real speech costly to collect and retain, synthetic speech from modern text-to-speech (TTS) is an appealing alternative for training automatic speech recognition (ASR) without exposing sensitive customer recordings. Yet a persistent distributional gap between synthetic and real data limits how far it can replace genuine recordings. Prior work largely treats this gap as a black box to be engineered around, but in our work, we instead examine its origin directly by probing a SLAM-ASR architecture. Then, we locali"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.29031","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/2606.29031/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":"2606.29031","created_at":"2026-06-30T01:17:50.091394+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.29031v1","created_at":"2026-06-30T01:17:50.091394+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.29031","created_at":"2026-06-30T01:17:50.091394+00:00"},{"alias_kind":"pith_short_12","alias_value":"4RH6VKKNWUKI","created_at":"2026-06-30T01:17:50.091394+00:00"},{"alias_kind":"pith_short_16","alias_value":"4RH6VKKNWUKIS7F6","created_at":"2026-06-30T01:17:50.091394+00:00"},{"alias_kind":"pith_short_8","alias_value":"4RH6VKKN","created_at":"2026-06-30T01:17:50.091394+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/4RH6VKKNWUKIS7F6O5I52DE77O","json":"https://pith.science/pith/4RH6VKKNWUKIS7F6O5I52DE77O.json","graph_json":"https://pith.science/api/pith-number/4RH6VKKNWUKIS7F6O5I52DE77O/graph.json","events_json":"https://pith.science/api/pith-number/4RH6VKKNWUKIS7F6O5I52DE77O/events.json","paper":"https://pith.science/paper/4RH6VKKN"},"agent_actions":{"view_html":"https://pith.science/pith/4RH6VKKNWUKIS7F6O5I52DE77O","download_json":"https://pith.science/pith/4RH6VKKNWUKIS7F6O5I52DE77O.json","view_paper":"https://pith.science/paper/4RH6VKKN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.29031&json=true","fetch_graph":"https://pith.science/api/pith-number/4RH6VKKNWUKIS7F6O5I52DE77O/graph.json","fetch_events":"https://pith.science/api/pith-number/4RH6VKKNWUKIS7F6O5I52DE77O/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4RH6VKKNWUKIS7F6O5I52DE77O/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4RH6VKKNWUKIS7F6O5I52DE77O/action/storage_attestation","attest_author":"https://pith.science/pith/4RH6VKKNWUKIS7F6O5I52DE77O/action/author_attestation","sign_citation":"https://pith.science/pith/4RH6VKKNWUKIS7F6O5I52DE77O/action/citation_signature","submit_replication":"https://pith.science/pith/4RH6VKKNWUKIS7F6O5I52DE77O/action/replication_record"}},"created_at":"2026-06-30T01:17:50.091394+00:00","updated_at":"2026-06-30T01:17:50.091394+00:00"}