{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:NU6JYNQO4BTR5MPCACKXNWQBUV","short_pith_number":"pith:NU6JYNQO","schema_version":"1.0","canonical_sha256":"6d3c9c360ee0671eb1e2009576da01a568126a926a7fd363d166a36bc29d1861","source":{"kind":"arxiv","id":"2104.03521","version":1},"attestation_state":"computed","paper":{"title":"Towards Multi-Scale Style Control for Expressive Speech Synthesis","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","eess.AS"],"primary_cat":"cs.SD","authors_text":"Changhe Song, Helen Meng, Jia Jia, Jingbei Li, Xiang Li, Zhiyong Wu","submitted_at":"2021-04-08T05:50:09Z","abstract_excerpt":"This paper introduces a multi-scale speech style modeling method for end-to-end expressive speech synthesis. The proposed method employs a multi-scale reference encoder to extract both the global-scale utterance-level and the local-scale quasi-phoneme-level style features of the target speech, which are then fed into the speech synthesis model as an extension to the input phoneme sequence. During training time, the multi-scale style model could be jointly trained with the speech synthesis model in an end-to-end fashion. By applying the proposed method to style transfer task, experimental resul"},"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":"2104.03521","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SD","submitted_at":"2021-04-08T05:50:09Z","cross_cats_sorted":["cs.CL","eess.AS"],"title_canon_sha256":"a0a7f0420bac60855f784a601bdd92e5f986348917294c89d309a1033f7e1400","abstract_canon_sha256":"c8119028b74b43fe46b5c68033984842baf4360eb1960a4645e0e09ab00aecf2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:30:18.380048Z","signature_b64":"756cmFrb7JB36NzADhjAx6gLjpudeJwdxTCKJCG+ej/d1WDLGEbrD6W7erFQKEMzOlBd7jNO9r6xY4wmhYlFBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6d3c9c360ee0671eb1e2009576da01a568126a926a7fd363d166a36bc29d1861","last_reissued_at":"2026-07-05T02:30:18.379574Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:30:18.379574Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards Multi-Scale Style Control for Expressive Speech Synthesis","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","eess.AS"],"primary_cat":"cs.SD","authors_text":"Changhe Song, Helen Meng, Jia Jia, Jingbei Li, Xiang Li, Zhiyong Wu","submitted_at":"2021-04-08T05:50:09Z","abstract_excerpt":"This paper introduces a multi-scale speech style modeling method for end-to-end expressive speech synthesis. The proposed method employs a multi-scale reference encoder to extract both the global-scale utterance-level and the local-scale quasi-phoneme-level style features of the target speech, which are then fed into the speech synthesis model as an extension to the input phoneme sequence. During training time, the multi-scale style model could be jointly trained with the speech synthesis model in an end-to-end fashion. By applying the proposed method to style transfer task, experimental resul"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2104.03521","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/2104.03521/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":"2104.03521","created_at":"2026-07-05T02:30:18.379630+00:00"},{"alias_kind":"arxiv_version","alias_value":"2104.03521v1","created_at":"2026-07-05T02:30:18.379630+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2104.03521","created_at":"2026-07-05T02:30:18.379630+00:00"},{"alias_kind":"pith_short_12","alias_value":"NU6JYNQO4BTR","created_at":"2026-07-05T02:30:18.379630+00:00"},{"alias_kind":"pith_short_16","alias_value":"NU6JYNQO4BTR5MPC","created_at":"2026-07-05T02:30:18.379630+00:00"},{"alias_kind":"pith_short_8","alias_value":"NU6JYNQO","created_at":"2026-07-05T02:30:18.379630+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.18748","citing_title":"Towards Expressive Video Dubbing with Multiscale Multimodal Context Interaction","ref_index":18,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NU6JYNQO4BTR5MPCACKXNWQBUV","json":"https://pith.science/pith/NU6JYNQO4BTR5MPCACKXNWQBUV.json","graph_json":"https://pith.science/api/pith-number/NU6JYNQO4BTR5MPCACKXNWQBUV/graph.json","events_json":"https://pith.science/api/pith-number/NU6JYNQO4BTR5MPCACKXNWQBUV/events.json","paper":"https://pith.science/paper/NU6JYNQO"},"agent_actions":{"view_html":"https://pith.science/pith/NU6JYNQO4BTR5MPCACKXNWQBUV","download_json":"https://pith.science/pith/NU6JYNQO4BTR5MPCACKXNWQBUV.json","view_paper":"https://pith.science/paper/NU6JYNQO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2104.03521&json=true","fetch_graph":"https://pith.science/api/pith-number/NU6JYNQO4BTR5MPCACKXNWQBUV/graph.json","fetch_events":"https://pith.science/api/pith-number/NU6JYNQO4BTR5MPCACKXNWQBUV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NU6JYNQO4BTR5MPCACKXNWQBUV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NU6JYNQO4BTR5MPCACKXNWQBUV/action/storage_attestation","attest_author":"https://pith.science/pith/NU6JYNQO4BTR5MPCACKXNWQBUV/action/author_attestation","sign_citation":"https://pith.science/pith/NU6JYNQO4BTR5MPCACKXNWQBUV/action/citation_signature","submit_replication":"https://pith.science/pith/NU6JYNQO4BTR5MPCACKXNWQBUV/action/replication_record"}},"created_at":"2026-07-05T02:30:18.379630+00:00","updated_at":"2026-07-05T02:30:18.379630+00:00"}