{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:OMTSAE5TS3NK37UHF7E4BMHM4L","short_pith_number":"pith:OMTSAE5T","schema_version":"1.0","canonical_sha256":"73272013b396daadfe872fc9c0b0ece2f37ade47cd017e456e3fb8cb180e1d55","source":{"kind":"arxiv","id":"2512.20978","version":2},"attestation_state":"computed","paper":{"title":"GenTSE: Enhancing Target Speaker Extraction via a Coarse-to-Fine Generative Language Model","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"eess.AS","authors_text":"Azmat Adnan, Boon Siew Han, Eng Siong Chng, Haoyang Li, Shreyas Gopal, Wei Rao, Xuyi Zhuang, Ye Ni, Yuanjin Zheng","submitted_at":"2025-12-24T06:13:02Z","abstract_excerpt":"Language Model (LM)-based generative modeling has emerged as a promising direction for TSE, offering potential for improved generalization and high-fidelity speech. We propose GenTSE, a two-stage decoder-only generative LM for TSE: Stage-1 predicts coarse semantic tokens, and Stage-2 generates fine acoustic tokens. Separating semantics and acoustics stabilizes decoding and yields more accurate target speech. Both stages use continuous SSL or codec embeddings, offering richer context than discretized-prompt methods. To reduce exposure bias, we employ a Frozen-LM Conditioning training strategy t"},"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":"2512.20978","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2025-12-24T06:13:02Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"053111ff082dac032858bb807a2e75556b19e95013c7bb4dcf354799b133e3dc","abstract_canon_sha256":"198ff9bce1c4fdc94f1f61a423255968cb8b0fa060d8a037e9b67ec34b8ddb4e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T01:05:12.170660Z","signature_b64":"Qbw1dimRxw6ekN9ms4IhEs1VqN40E9NTB1McPEtK/l2CAtfbEKtu1mgcOgwwUjCWpxCzAz9aNKkm61Ni3wCdBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"73272013b396daadfe872fc9c0b0ece2f37ade47cd017e456e3fb8cb180e1d55","last_reissued_at":"2026-06-09T01:05:12.170145Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T01:05:12.170145Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"GenTSE: Enhancing Target Speaker Extraction via a Coarse-to-Fine Generative Language Model","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"eess.AS","authors_text":"Azmat Adnan, Boon Siew Han, Eng Siong Chng, Haoyang Li, Shreyas Gopal, Wei Rao, Xuyi Zhuang, Ye Ni, Yuanjin Zheng","submitted_at":"2025-12-24T06:13:02Z","abstract_excerpt":"Language Model (LM)-based generative modeling has emerged as a promising direction for TSE, offering potential for improved generalization and high-fidelity speech. We propose GenTSE, a two-stage decoder-only generative LM for TSE: Stage-1 predicts coarse semantic tokens, and Stage-2 generates fine acoustic tokens. Separating semantics and acoustics stabilizes decoding and yields more accurate target speech. Both stages use continuous SSL or codec embeddings, offering richer context than discretized-prompt methods. To reduce exposure bias, we employ a Frozen-LM Conditioning training strategy t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2512.20978","kind":"arxiv","version":2},"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/2512.20978/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":"2512.20978","created_at":"2026-06-09T01:05:12.170209+00:00"},{"alias_kind":"arxiv_version","alias_value":"2512.20978v2","created_at":"2026-06-09T01:05:12.170209+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.20978","created_at":"2026-06-09T01:05:12.170209+00:00"},{"alias_kind":"pith_short_12","alias_value":"OMTSAE5TS3NK","created_at":"2026-06-09T01:05:12.170209+00:00"},{"alias_kind":"pith_short_16","alias_value":"OMTSAE5TS3NK37UH","created_at":"2026-06-09T01:05:12.170209+00:00"},{"alias_kind":"pith_short_8","alias_value":"OMTSAE5T","created_at":"2026-06-09T01:05:12.170209+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/OMTSAE5TS3NK37UHF7E4BMHM4L","json":"https://pith.science/pith/OMTSAE5TS3NK37UHF7E4BMHM4L.json","graph_json":"https://pith.science/api/pith-number/OMTSAE5TS3NK37UHF7E4BMHM4L/graph.json","events_json":"https://pith.science/api/pith-number/OMTSAE5TS3NK37UHF7E4BMHM4L/events.json","paper":"https://pith.science/paper/OMTSAE5T"},"agent_actions":{"view_html":"https://pith.science/pith/OMTSAE5TS3NK37UHF7E4BMHM4L","download_json":"https://pith.science/pith/OMTSAE5TS3NK37UHF7E4BMHM4L.json","view_paper":"https://pith.science/paper/OMTSAE5T","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2512.20978&json=true","fetch_graph":"https://pith.science/api/pith-number/OMTSAE5TS3NK37UHF7E4BMHM4L/graph.json","fetch_events":"https://pith.science/api/pith-number/OMTSAE5TS3NK37UHF7E4BMHM4L/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OMTSAE5TS3NK37UHF7E4BMHM4L/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OMTSAE5TS3NK37UHF7E4BMHM4L/action/storage_attestation","attest_author":"https://pith.science/pith/OMTSAE5TS3NK37UHF7E4BMHM4L/action/author_attestation","sign_citation":"https://pith.science/pith/OMTSAE5TS3NK37UHF7E4BMHM4L/action/citation_signature","submit_replication":"https://pith.science/pith/OMTSAE5TS3NK37UHF7E4BMHM4L/action/replication_record"}},"created_at":"2026-06-09T01:05:12.170209+00:00","updated_at":"2026-06-09T01:05:12.170209+00:00"}