{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:B3MSYYNUE7XBZNQSWT4KHFWKGE","short_pith_number":"pith:B3MSYYNU","schema_version":"1.0","canonical_sha256":"0ed92c61b427ee1cb612b4f8a396ca31291c9aede7421e9e9d74aa75c7cf2ed5","source":{"kind":"arxiv","id":"2502.00377","version":1},"attestation_state":"computed","paper":{"title":"When End-to-End is Overkill: Rethinking Cascaded Speech-to-Text Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.MM","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Anna Min, Chenxu Hu, Hang Zhao, Yi Ren","submitted_at":"2025-02-01T09:29:21Z","abstract_excerpt":"Though end-to-end speech-to-text translation has been a great success, we argue that the cascaded speech-to-text translation model still has its place, which is usually criticized for the error propagation between automatic speech recognition (ASR) and machine translation (MT) models. In this paper, we explore the benefits of incorporating multiple candidates from ASR and self-supervised speech features into MT. Our analysis reveals that the primary cause of cascading errors stems from the increased divergence between similar samples in the speech domain when mapped to the text domain. By incl"},"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":"2502.00377","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-02-01T09:29:21Z","cross_cats_sorted":["cs.AI","cs.MM","cs.SD","eess.AS"],"title_canon_sha256":"93de35926995c4a98de561507d22d2e5ec5f5ef4d7aa144fc0b2e3f497a86fd6","abstract_canon_sha256":"3dd53db295f5e837fd7ba37d13e8d29a9ccf3df6229599b68b55ce5536029844"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:08:32.881938Z","signature_b64":"2FDrx5b5fwJOrGq66+TUWpxGkLHJ0hqGefTZtZhIfGChuRg3DaomI8bWw9Nlo74SSQuzlU0NUhq5fOyM78ohAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0ed92c61b427ee1cb612b4f8a396ca31291c9aede7421e9e9d74aa75c7cf2ed5","last_reissued_at":"2026-07-05T10:08:32.881536Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:08:32.881536Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"When End-to-End is Overkill: Rethinking Cascaded Speech-to-Text Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.MM","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Anna Min, Chenxu Hu, Hang Zhao, Yi Ren","submitted_at":"2025-02-01T09:29:21Z","abstract_excerpt":"Though end-to-end speech-to-text translation has been a great success, we argue that the cascaded speech-to-text translation model still has its place, which is usually criticized for the error propagation between automatic speech recognition (ASR) and machine translation (MT) models. In this paper, we explore the benefits of incorporating multiple candidates from ASR and self-supervised speech features into MT. Our analysis reveals that the primary cause of cascading errors stems from the increased divergence between similar samples in the speech domain when mapped to the text domain. By incl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.00377","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/2502.00377/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":"2502.00377","created_at":"2026-07-05T10:08:32.881591+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.00377v1","created_at":"2026-07-05T10:08:32.881591+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.00377","created_at":"2026-07-05T10:08:32.881591+00:00"},{"alias_kind":"pith_short_12","alias_value":"B3MSYYNUE7XB","created_at":"2026-07-05T10:08:32.881591+00:00"},{"alias_kind":"pith_short_16","alias_value":"B3MSYYNUE7XBZNQS","created_at":"2026-07-05T10:08:32.881591+00:00"},{"alias_kind":"pith_short_8","alias_value":"B3MSYYNU","created_at":"2026-07-05T10:08:32.881591+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2512.16378","citing_title":"Hearing to Translate: The Effectiveness of Speech Modality Integration into LLMs","ref_index":62,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/B3MSYYNUE7XBZNQSWT4KHFWKGE","json":"https://pith.science/pith/B3MSYYNUE7XBZNQSWT4KHFWKGE.json","graph_json":"https://pith.science/api/pith-number/B3MSYYNUE7XBZNQSWT4KHFWKGE/graph.json","events_json":"https://pith.science/api/pith-number/B3MSYYNUE7XBZNQSWT4KHFWKGE/events.json","paper":"https://pith.science/paper/B3MSYYNU"},"agent_actions":{"view_html":"https://pith.science/pith/B3MSYYNUE7XBZNQSWT4KHFWKGE","download_json":"https://pith.science/pith/B3MSYYNUE7XBZNQSWT4KHFWKGE.json","view_paper":"https://pith.science/paper/B3MSYYNU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.00377&json=true","fetch_graph":"https://pith.science/api/pith-number/B3MSYYNUE7XBZNQSWT4KHFWKGE/graph.json","fetch_events":"https://pith.science/api/pith-number/B3MSYYNUE7XBZNQSWT4KHFWKGE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/B3MSYYNUE7XBZNQSWT4KHFWKGE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/B3MSYYNUE7XBZNQSWT4KHFWKGE/action/storage_attestation","attest_author":"https://pith.science/pith/B3MSYYNUE7XBZNQSWT4KHFWKGE/action/author_attestation","sign_citation":"https://pith.science/pith/B3MSYYNUE7XBZNQSWT4KHFWKGE/action/citation_signature","submit_replication":"https://pith.science/pith/B3MSYYNUE7XBZNQSWT4KHFWKGE/action/replication_record"}},"created_at":"2026-07-05T10:08:32.881591+00:00","updated_at":"2026-07-05T10:08:32.881591+00:00"}