{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:ET2FI4ZPS5T7QUWMXPEF33H4GF","short_pith_number":"pith:ET2FI4ZP","schema_version":"1.0","canonical_sha256":"24f454732f9767f852ccbbc85decfc314dc0bce9e91b0ec7de17a1683f2540f0","source":{"kind":"arxiv","id":"2010.10504","version":2},"attestation_state":"computed","paper":{"title":"Pushing the Limits of Semi-Supervised Learning for Automatic Speech Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Chung-Cheng Chiu, Daniel S. Park, James Qin, Quoc V. Le, Ruoming Pang, Wei Han, Yonghui Wu, Yu Zhang","submitted_at":"2020-10-20T17:58:13Z","abstract_excerpt":"We employ a combination of recent developments in semi-supervised learning for automatic speech recognition to obtain state-of-the-art results on LibriSpeech utilizing the unlabeled audio of the Libri-Light dataset. More precisely, we carry out noisy student training with SpecAugment using giant Conformer models pre-trained using wav2vec 2.0 pre-training. By doing so, we are able to achieve word-error-rates (WERs) 1.4%/2.6% on the LibriSpeech test/test-other sets against the current state-of-the-art WERs 1.7%/3.3%."},"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":"2010.10504","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2020-10-20T17:58:13Z","cross_cats_sorted":["cs.LG","cs.SD"],"title_canon_sha256":"c6dc028788689e4130977ed44585559ea35bf316b5e6f0dbda3cd26a8226c934","abstract_canon_sha256":"fa51d741d229eee8c731ad5c23c8108c911f810078b25f9c9d52c45888cea9be"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:42:12.907060Z","signature_b64":"Tm8vn8jHViOwUs5v3IN92vorVjJ7uq46cYky8Q2QphTssnhLsmC84pnbUl0AJs3PCZ8hGdZhxY8ZLs/5k1niDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"24f454732f9767f852ccbbc85decfc314dc0bce9e91b0ec7de17a1683f2540f0","last_reissued_at":"2026-07-05T04:42:12.906627Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:42:12.906627Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Pushing the Limits of Semi-Supervised Learning for Automatic Speech Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Chung-Cheng Chiu, Daniel S. Park, James Qin, Quoc V. Le, Ruoming Pang, Wei Han, Yonghui Wu, Yu Zhang","submitted_at":"2020-10-20T17:58:13Z","abstract_excerpt":"We employ a combination of recent developments in semi-supervised learning for automatic speech recognition to obtain state-of-the-art results on LibriSpeech utilizing the unlabeled audio of the Libri-Light dataset. More precisely, we carry out noisy student training with SpecAugment using giant Conformer models pre-trained using wav2vec 2.0 pre-training. By doing so, we are able to achieve word-error-rates (WERs) 1.4%/2.6% on the LibriSpeech test/test-other sets against the current state-of-the-art WERs 1.7%/3.3%."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.10504","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/2010.10504/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":"2010.10504","created_at":"2026-07-05T04:42:12.906683+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.10504v2","created_at":"2026-07-05T04:42:12.906683+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.10504","created_at":"2026-07-05T04:42:12.906683+00:00"},{"alias_kind":"pith_short_12","alias_value":"ET2FI4ZPS5T7","created_at":"2026-07-05T04:42:12.906683+00:00"},{"alias_kind":"pith_short_16","alias_value":"ET2FI4ZPS5T7QUWM","created_at":"2026-07-05T04:42:12.906683+00:00"},{"alias_kind":"pith_short_8","alias_value":"ET2FI4ZP","created_at":"2026-07-05T04:42:12.906683+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.18659","citing_title":"Responsible ASR: Overcoming Challenges of Foundational Models in Narrow-Band and Low-Resource Settings","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10360","citing_title":"ViP-VL: Vietnamese Self-supervised Speech Pretraining Model with Vector-Quantization Learning","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2309.16588","citing_title":"Vision Transformers Need Registers","ref_index":135,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23077","citing_title":"Adopting State-of-the-Art Pretrained Audio Representations for Music Recommender Systems","ref_index":104,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ET2FI4ZPS5T7QUWMXPEF33H4GF","json":"https://pith.science/pith/ET2FI4ZPS5T7QUWMXPEF33H4GF.json","graph_json":"https://pith.science/api/pith-number/ET2FI4ZPS5T7QUWMXPEF33H4GF/graph.json","events_json":"https://pith.science/api/pith-number/ET2FI4ZPS5T7QUWMXPEF33H4GF/events.json","paper":"https://pith.science/paper/ET2FI4ZP"},"agent_actions":{"view_html":"https://pith.science/pith/ET2FI4ZPS5T7QUWMXPEF33H4GF","download_json":"https://pith.science/pith/ET2FI4ZPS5T7QUWMXPEF33H4GF.json","view_paper":"https://pith.science/paper/ET2FI4ZP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.10504&json=true","fetch_graph":"https://pith.science/api/pith-number/ET2FI4ZPS5T7QUWMXPEF33H4GF/graph.json","fetch_events":"https://pith.science/api/pith-number/ET2FI4ZPS5T7QUWMXPEF33H4GF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ET2FI4ZPS5T7QUWMXPEF33H4GF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ET2FI4ZPS5T7QUWMXPEF33H4GF/action/storage_attestation","attest_author":"https://pith.science/pith/ET2FI4ZPS5T7QUWMXPEF33H4GF/action/author_attestation","sign_citation":"https://pith.science/pith/ET2FI4ZPS5T7QUWMXPEF33H4GF/action/citation_signature","submit_replication":"https://pith.science/pith/ET2FI4ZPS5T7QUWMXPEF33H4GF/action/replication_record"}},"created_at":"2026-07-05T04:42:12.906683+00:00","updated_at":"2026-07-05T04:42:12.906683+00:00"}