{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:TYDUPK6V7JBV7WI4FEISXQ2URK","short_pith_number":"pith:TYDUPK6V","schema_version":"1.0","canonical_sha256":"9e0747abd5fa435fd91c29112bc3548a8637c8392562ce3f27d1ad7e453f2aa5","source":{"kind":"arxiv","id":"2505.24200","version":2},"attestation_state":"computed","paper":{"title":"Improving Multilingual Speech Models on ML-SUPERB 2.0: Fine-tuning with Data Augmentation and LID-Aware CTC","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","eess.AS"],"primary_cat":"cs.SD","authors_text":"Jiancheng Sun, Qingzheng Wang, Shinji Watanabe, Yifan Peng","submitted_at":"2025-05-30T04:25:15Z","abstract_excerpt":"Multilingual speech processing with self-supervised or supervised pre-trained Speech Foundation Models (SFM) has achieved strong performance on tasks like Language Identification (LID) and Automatic Speech Recognition (ASR). However, these models struggle with limited resources during fine-tuning. This paper enhances multilingual LID and ASR on ML-SUPERB 2.0 by exploring multiple strategies for adapting SFMs, including frozen upstream training, partial fine-tuning, and low-rank adaptation. Furthermore, we employ data augmentation to mitigate performance gaps in few-shot settings and introduce "},"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":"2505.24200","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SD","submitted_at":"2025-05-30T04:25:15Z","cross_cats_sorted":["cs.CL","eess.AS"],"title_canon_sha256":"41271bc26eac34ba8092c231b373ae73be057054c103f4923c3a74368ed63714","abstract_canon_sha256":"fb93e1ea277d404f4dcdc93e54494fe02c5bd78342dd40b71ae7296d08bd4eb1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:15:01.072309Z","signature_b64":"67dA8fpnIx9il9abxEBcRsHiYarP3e14/rcr4YfToXvSyVGRotnAPr1H6Z4uTHVTi5wLXptRQHQ+ihqR5jySDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9e0747abd5fa435fd91c29112bc3548a8637c8392562ce3f27d1ad7e453f2aa5","last_reissued_at":"2026-07-05T11:15:01.071824Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:15:01.071824Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Improving Multilingual Speech Models on ML-SUPERB 2.0: Fine-tuning with Data Augmentation and LID-Aware CTC","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","eess.AS"],"primary_cat":"cs.SD","authors_text":"Jiancheng Sun, Qingzheng Wang, Shinji Watanabe, Yifan Peng","submitted_at":"2025-05-30T04:25:15Z","abstract_excerpt":"Multilingual speech processing with self-supervised or supervised pre-trained Speech Foundation Models (SFM) has achieved strong performance on tasks like Language Identification (LID) and Automatic Speech Recognition (ASR). However, these models struggle with limited resources during fine-tuning. This paper enhances multilingual LID and ASR on ML-SUPERB 2.0 by exploring multiple strategies for adapting SFMs, including frozen upstream training, partial fine-tuning, and low-rank adaptation. Furthermore, we employ data augmentation to mitigate performance gaps in few-shot settings and introduce "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.24200","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/2505.24200/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":"2505.24200","created_at":"2026-07-05T11:15:01.071883+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.24200v2","created_at":"2026-07-05T11:15:01.071883+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.24200","created_at":"2026-07-05T11:15:01.071883+00:00"},{"alias_kind":"pith_short_12","alias_value":"TYDUPK6V7JBV","created_at":"2026-07-05T11:15:01.071883+00:00"},{"alias_kind":"pith_short_16","alias_value":"TYDUPK6V7JBV7WI4","created_at":"2026-07-05T11:15:01.071883+00:00"},{"alias_kind":"pith_short_8","alias_value":"TYDUPK6V","created_at":"2026-07-05T11:15:01.071883+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.24200","citing_title":"Improving Multilingual Speech Models on ML-SUPERB 2.0: Fine-tuning with Data Augmentation and LID-Aware CTC","ref_index":3,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TYDUPK6V7JBV7WI4FEISXQ2URK","json":"https://pith.science/pith/TYDUPK6V7JBV7WI4FEISXQ2URK.json","graph_json":"https://pith.science/api/pith-number/TYDUPK6V7JBV7WI4FEISXQ2URK/graph.json","events_json":"https://pith.science/api/pith-number/TYDUPK6V7JBV7WI4FEISXQ2URK/events.json","paper":"https://pith.science/paper/TYDUPK6V"},"agent_actions":{"view_html":"https://pith.science/pith/TYDUPK6V7JBV7WI4FEISXQ2URK","download_json":"https://pith.science/pith/TYDUPK6V7JBV7WI4FEISXQ2URK.json","view_paper":"https://pith.science/paper/TYDUPK6V","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.24200&json=true","fetch_graph":"https://pith.science/api/pith-number/TYDUPK6V7JBV7WI4FEISXQ2URK/graph.json","fetch_events":"https://pith.science/api/pith-number/TYDUPK6V7JBV7WI4FEISXQ2URK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TYDUPK6V7JBV7WI4FEISXQ2URK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TYDUPK6V7JBV7WI4FEISXQ2URK/action/storage_attestation","attest_author":"https://pith.science/pith/TYDUPK6V7JBV7WI4FEISXQ2URK/action/author_attestation","sign_citation":"https://pith.science/pith/TYDUPK6V7JBV7WI4FEISXQ2URK/action/citation_signature","submit_replication":"https://pith.science/pith/TYDUPK6V7JBV7WI4FEISXQ2URK/action/replication_record"}},"created_at":"2026-07-05T11:15:01.071883+00:00","updated_at":"2026-07-05T11:15:01.071883+00:00"}