{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:IOKQW7JRT7IQXZFHAKLUCEQSVQ","short_pith_number":"pith:IOKQW7JR","schema_version":"1.0","canonical_sha256":"43950b7d319fd10be4a70297411212ac2fffd70dd6eeac40c070acddc34382a8","source":{"kind":"arxiv","id":"2006.02490","version":2},"attestation_state":"computed","paper":{"title":"Self-Training for End-to-End Speech Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Juan Pino, Mohammad Javad Dousti, Qiantong Xu, Xutai Ma, Yun Tang","submitted_at":"2020-06-03T19:28:36Z","abstract_excerpt":"One of the main challenges for end-to-end speech translation is data scarcity. We leverage pseudo-labels generated from unlabeled audio by a cascade and an end-to-end speech translation model. This provides 8.3 and 5.7 BLEU gains over a strong semi-supervised baseline on the MuST-C English-French and English-German datasets, reaching state-of-the art performance. The effect of the quality of the pseudo-labels is investigated. Our approach is shown to be more effective than simply pre-training the encoder on the speech recognition task. Finally, we demonstrate the effectiveness of self-training"},"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":"2006.02490","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-06-03T19:28:36Z","cross_cats_sorted":["cs.SD","eess.AS"],"title_canon_sha256":"1edaaf5a83a991bb5551bee9105b68b0bde8d492248066832e852d620a09709f","abstract_canon_sha256":"e42c1d1efacbca6edbbcaab4112b41030e8ed36df05a2d6b7c482a5a5ab3fc8e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:42:14.249621Z","signature_b64":"qE9+2PIfjsDtBlRpoXeZlhXrxmZE3J5ctG2Mx06nryLjrADU3DmuAmWZBC2rAz6OTenpiGeM8I0p0RJJXgkZDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"43950b7d319fd10be4a70297411212ac2fffd70dd6eeac40c070acddc34382a8","last_reissued_at":"2026-07-05T01:42:14.249227Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:42:14.249227Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Self-Training for End-to-End Speech Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Juan Pino, Mohammad Javad Dousti, Qiantong Xu, Xutai Ma, Yun Tang","submitted_at":"2020-06-03T19:28:36Z","abstract_excerpt":"One of the main challenges for end-to-end speech translation is data scarcity. We leverage pseudo-labels generated from unlabeled audio by a cascade and an end-to-end speech translation model. This provides 8.3 and 5.7 BLEU gains over a strong semi-supervised baseline on the MuST-C English-French and English-German datasets, reaching state-of-the art performance. The effect of the quality of the pseudo-labels is investigated. Our approach is shown to be more effective than simply pre-training the encoder on the speech recognition task. Finally, we demonstrate the effectiveness of self-training"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2006.02490","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/2006.02490/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":"2006.02490","created_at":"2026-07-05T01:42:14.249285+00:00"},{"alias_kind":"arxiv_version","alias_value":"2006.02490v2","created_at":"2026-07-05T01:42:14.249285+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2006.02490","created_at":"2026-07-05T01:42:14.249285+00:00"},{"alias_kind":"pith_short_12","alias_value":"IOKQW7JRT7IQ","created_at":"2026-07-05T01:42:14.249285+00:00"},{"alias_kind":"pith_short_16","alias_value":"IOKQW7JRT7IQXZFH","created_at":"2026-07-05T01:42:14.249285+00:00"},{"alias_kind":"pith_short_8","alias_value":"IOKQW7JR","created_at":"2026-07-05T01:42:14.249285+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.00377","citing_title":"When End-to-End is Overkill: Rethinking Cascaded Speech-to-Text Translation","ref_index":5,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IOKQW7JRT7IQXZFHAKLUCEQSVQ","json":"https://pith.science/pith/IOKQW7JRT7IQXZFHAKLUCEQSVQ.json","graph_json":"https://pith.science/api/pith-number/IOKQW7JRT7IQXZFHAKLUCEQSVQ/graph.json","events_json":"https://pith.science/api/pith-number/IOKQW7JRT7IQXZFHAKLUCEQSVQ/events.json","paper":"https://pith.science/paper/IOKQW7JR"},"agent_actions":{"view_html":"https://pith.science/pith/IOKQW7JRT7IQXZFHAKLUCEQSVQ","download_json":"https://pith.science/pith/IOKQW7JRT7IQXZFHAKLUCEQSVQ.json","view_paper":"https://pith.science/paper/IOKQW7JR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2006.02490&json=true","fetch_graph":"https://pith.science/api/pith-number/IOKQW7JRT7IQXZFHAKLUCEQSVQ/graph.json","fetch_events":"https://pith.science/api/pith-number/IOKQW7JRT7IQXZFHAKLUCEQSVQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IOKQW7JRT7IQXZFHAKLUCEQSVQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IOKQW7JRT7IQXZFHAKLUCEQSVQ/action/storage_attestation","attest_author":"https://pith.science/pith/IOKQW7JRT7IQXZFHAKLUCEQSVQ/action/author_attestation","sign_citation":"https://pith.science/pith/IOKQW7JRT7IQXZFHAKLUCEQSVQ/action/citation_signature","submit_replication":"https://pith.science/pith/IOKQW7JRT7IQXZFHAKLUCEQSVQ/action/replication_record"}},"created_at":"2026-07-05T01:42:14.249285+00:00","updated_at":"2026-07-05T01:42:14.249285+00:00"}