{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:HDY7ZZOWRKO2DCSEKL5BA5UE4Q","short_pith_number":"pith:HDY7ZZOW","schema_version":"1.0","canonical_sha256":"38f1fce5d68a9da18a4452fa107684e43396af8257f825c12ff3a27b33ce9db5","source":{"kind":"arxiv","id":"2108.10447","version":1},"attestation_state":"computed","paper":{"title":"One TTS Alignment To Rule Them All","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG","eess.AS"],"primary_cat":"cs.SD","authors_text":"Adrian {\\L}ancucki, Bryan Catanzaro, Kevin J. Shih, Rafael Valle, Rohan Badlani, Wei Ping","submitted_at":"2021-08-23T23:45:48Z","abstract_excerpt":"Speech-to-text alignment is a critical component of neural textto-speech (TTS) models. Autoregressive TTS models typically use an attention mechanism to learn these alignments on-line. However, these alignments tend to be brittle and often fail to generalize to long utterances and out-of-domain text, leading to missing or repeating words. Most non-autoregressive endto-end TTS models rely on durations extracted from external sources. In this paper we leverage the alignment mechanism proposed in RAD-TTS as a generic alignment learning framework, easily applicable to a variety of neural TTS model"},"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":"2108.10447","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SD","submitted_at":"2021-08-23T23:45:48Z","cross_cats_sorted":["cs.CL","cs.LG","eess.AS"],"title_canon_sha256":"e67b1b4d379603ba35a3107bf7dd4b103cb21339be628d464d39a313430a22bb","abstract_canon_sha256":"0b71458a97c8419dd6345500f0cd1a5eb319189819bc967accd5cfe17817eeec"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:08:09.740352Z","signature_b64":"dsxougYTRfPrv8YArjbm+UX5WWHeTXyNUEYhQYK/b8ZyE+UJwADUAVRlzvgemt2792qiQsqDRBZKwg15moMiDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"38f1fce5d68a9da18a4452fa107684e43396af8257f825c12ff3a27b33ce9db5","last_reissued_at":"2026-07-05T03:08:09.740000Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:08:09.740000Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"One TTS Alignment To Rule Them All","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG","eess.AS"],"primary_cat":"cs.SD","authors_text":"Adrian {\\L}ancucki, Bryan Catanzaro, Kevin J. Shih, Rafael Valle, Rohan Badlani, Wei Ping","submitted_at":"2021-08-23T23:45:48Z","abstract_excerpt":"Speech-to-text alignment is a critical component of neural textto-speech (TTS) models. Autoregressive TTS models typically use an attention mechanism to learn these alignments on-line. However, these alignments tend to be brittle and often fail to generalize to long utterances and out-of-domain text, leading to missing or repeating words. Most non-autoregressive endto-end TTS models rely on durations extracted from external sources. In this paper we leverage the alignment mechanism proposed in RAD-TTS as a generic alignment learning framework, easily applicable to a variety of neural TTS model"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2108.10447","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/2108.10447/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":"2108.10447","created_at":"2026-07-05T03:08:09.740056+00:00"},{"alias_kind":"arxiv_version","alias_value":"2108.10447v1","created_at":"2026-07-05T03:08:09.740056+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2108.10447","created_at":"2026-07-05T03:08:09.740056+00:00"},{"alias_kind":"pith_short_12","alias_value":"HDY7ZZOWRKO2","created_at":"2026-07-05T03:08:09.740056+00:00"},{"alias_kind":"pith_short_16","alias_value":"HDY7ZZOWRKO2DCSE","created_at":"2026-07-05T03:08:09.740056+00:00"},{"alias_kind":"pith_short_8","alias_value":"HDY7ZZOW","created_at":"2026-07-05T03:08:09.740056+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.01719","citing_title":"MJ-VIDEO: Fine-Grained Benchmarking and Rewarding Video Preferences in Video Generation","ref_index":2021,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HDY7ZZOWRKO2DCSEKL5BA5UE4Q","json":"https://pith.science/pith/HDY7ZZOWRKO2DCSEKL5BA5UE4Q.json","graph_json":"https://pith.science/api/pith-number/HDY7ZZOWRKO2DCSEKL5BA5UE4Q/graph.json","events_json":"https://pith.science/api/pith-number/HDY7ZZOWRKO2DCSEKL5BA5UE4Q/events.json","paper":"https://pith.science/paper/HDY7ZZOW"},"agent_actions":{"view_html":"https://pith.science/pith/HDY7ZZOWRKO2DCSEKL5BA5UE4Q","download_json":"https://pith.science/pith/HDY7ZZOWRKO2DCSEKL5BA5UE4Q.json","view_paper":"https://pith.science/paper/HDY7ZZOW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2108.10447&json=true","fetch_graph":"https://pith.science/api/pith-number/HDY7ZZOWRKO2DCSEKL5BA5UE4Q/graph.json","fetch_events":"https://pith.science/api/pith-number/HDY7ZZOWRKO2DCSEKL5BA5UE4Q/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HDY7ZZOWRKO2DCSEKL5BA5UE4Q/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HDY7ZZOWRKO2DCSEKL5BA5UE4Q/action/storage_attestation","attest_author":"https://pith.science/pith/HDY7ZZOWRKO2DCSEKL5BA5UE4Q/action/author_attestation","sign_citation":"https://pith.science/pith/HDY7ZZOWRKO2DCSEKL5BA5UE4Q/action/citation_signature","submit_replication":"https://pith.science/pith/HDY7ZZOWRKO2DCSEKL5BA5UE4Q/action/replication_record"}},"created_at":"2026-07-05T03:08:09.740056+00:00","updated_at":"2026-07-05T03:08:09.740056+00:00"}