{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:RTVADOAUQYHIU3IRYSOB6HBLMM","short_pith_number":"pith:RTVADOAU","schema_version":"1.0","canonical_sha256":"8cea01b814860e8a6d11c49c1f1c2b63394fe2b1204732d5881ddbb31f869e46","source":{"kind":"arxiv","id":"2606.23190","version":1},"attestation_state":"computed","paper":{"title":"FlowTTS-GRPO: Online Reinforcement Learning with Multi-Objective Reward Optimization for Flow-Matching Based Text-to-Speech","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SD"],"primary_cat":"eess.AS","authors_text":"Biao Tian, Han Zhao, Haoxu Wang, Weiqing Li, Xiangang Li, Xiang Lv","submitted_at":"2026-06-22T11:37:45Z","abstract_excerpt":"Existing Reinforcement Learning (RL) research for Text-to-Speech (TTS) focuses on large language models (LLMs), leaving Flow-Matching (FM) under-explored. We present FlowTTS-GRPO, an online RL framework for FM-based TTS. By converting ordinary differential equation (ODE) trajectories into stochastic differential equation (SDE) paths, our method enables direct fine-tuning of open-source FM models without auxiliary models. We show that a weighted reward combination converges faster than a probabilistic scheme, and identify three practical optimizations: omitting classifier-free guidance (CFG) du"},"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":"2606.23190","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2026-06-22T11:37:45Z","cross_cats_sorted":["cs.SD"],"title_canon_sha256":"c7baadfeb4b29d87016833f3647b6221a1606735d1c004a6a20b69a76841fa77","abstract_canon_sha256":"1faf2b5744876b278697b03ede276bf9c6dfe334a8a0c47469dd2968f5dc203c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T03:14:12.558998Z","signature_b64":"q6dq4PuG+vJx30WPk2fhh9tJK+/nh5rbHZFhUYf1G+IrP8X2H0KXNZZlv3U3IUWdrCcIpuHUMOQKBOyXcCO4AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8cea01b814860e8a6d11c49c1f1c2b63394fe2b1204732d5881ddbb31f869e46","last_reissued_at":"2026-06-23T03:14:12.558579Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T03:14:12.558579Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"FlowTTS-GRPO: Online Reinforcement Learning with Multi-Objective Reward Optimization for Flow-Matching Based Text-to-Speech","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SD"],"primary_cat":"eess.AS","authors_text":"Biao Tian, Han Zhao, Haoxu Wang, Weiqing Li, Xiangang Li, Xiang Lv","submitted_at":"2026-06-22T11:37:45Z","abstract_excerpt":"Existing Reinforcement Learning (RL) research for Text-to-Speech (TTS) focuses on large language models (LLMs), leaving Flow-Matching (FM) under-explored. We present FlowTTS-GRPO, an online RL framework for FM-based TTS. By converting ordinary differential equation (ODE) trajectories into stochastic differential equation (SDE) paths, our method enables direct fine-tuning of open-source FM models without auxiliary models. We show that a weighted reward combination converges faster than a probabilistic scheme, and identify three practical optimizations: omitting classifier-free guidance (CFG) du"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.23190","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/2606.23190/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":"2606.23190","created_at":"2026-06-23T03:14:12.558637+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.23190v1","created_at":"2026-06-23T03:14:12.558637+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.23190","created_at":"2026-06-23T03:14:12.558637+00:00"},{"alias_kind":"pith_short_12","alias_value":"RTVADOAUQYHI","created_at":"2026-06-23T03:14:12.558637+00:00"},{"alias_kind":"pith_short_16","alias_value":"RTVADOAUQYHIU3IR","created_at":"2026-06-23T03:14:12.558637+00:00"},{"alias_kind":"pith_short_8","alias_value":"RTVADOAU","created_at":"2026-06-23T03:14:12.558637+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2606.23190","citing_title":"FlowTTS-GRPO: Online Reinforcement Learning with Multi-Objective Reward Optimization for Flow-Matching Based Text-to-Speech","ref_index":1,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RTVADOAUQYHIU3IRYSOB6HBLMM","json":"https://pith.science/pith/RTVADOAUQYHIU3IRYSOB6HBLMM.json","graph_json":"https://pith.science/api/pith-number/RTVADOAUQYHIU3IRYSOB6HBLMM/graph.json","events_json":"https://pith.science/api/pith-number/RTVADOAUQYHIU3IRYSOB6HBLMM/events.json","paper":"https://pith.science/paper/RTVADOAU"},"agent_actions":{"view_html":"https://pith.science/pith/RTVADOAUQYHIU3IRYSOB6HBLMM","download_json":"https://pith.science/pith/RTVADOAUQYHIU3IRYSOB6HBLMM.json","view_paper":"https://pith.science/paper/RTVADOAU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.23190&json=true","fetch_graph":"https://pith.science/api/pith-number/RTVADOAUQYHIU3IRYSOB6HBLMM/graph.json","fetch_events":"https://pith.science/api/pith-number/RTVADOAUQYHIU3IRYSOB6HBLMM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RTVADOAUQYHIU3IRYSOB6HBLMM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RTVADOAUQYHIU3IRYSOB6HBLMM/action/storage_attestation","attest_author":"https://pith.science/pith/RTVADOAUQYHIU3IRYSOB6HBLMM/action/author_attestation","sign_citation":"https://pith.science/pith/RTVADOAUQYHIU3IRYSOB6HBLMM/action/citation_signature","submit_replication":"https://pith.science/pith/RTVADOAUQYHIU3IRYSOB6HBLMM/action/replication_record"}},"created_at":"2026-06-23T03:14:12.558637+00:00","updated_at":"2026-06-23T03:14:12.558637+00:00"}