{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:AC3QYYT5CYRVQGE5DDNCGSLWIC","short_pith_number":"pith:AC3QYYT5","schema_version":"1.0","canonical_sha256":"00b70c627d162358189d18da23497640a1576d3d26111c59e2037d2e627df94e","source":{"kind":"arxiv","id":"2502.01547","version":3},"attestation_state":"computed","paper":{"title":"mWhisper-Flamingo for Multilingual Audio-Visual Noise-Robust Speech Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.SD"],"primary_cat":"eess.AS","authors_text":"Andrew Rouditchenko, Hilde Kuehne, James Glass, Rogerio Feris, Samuel Thomas","submitted_at":"2025-02-03T17:29:52Z","abstract_excerpt":"Audio-Visual Speech Recognition (AVSR) combines lip-based video with audio and can improve performance in noise, but most methods are trained only on English data. One limitation is the lack of large-scale multilingual video data, which makes it hard to train models from scratch. In this work, we propose mWhisper-Flamingo for multilingual AVSR which combines the strengths of a pre-trained audio model (Whisper) and video model (AV-HuBERT). To enable better multi-modal integration and improve the noisy multilingual performance, we introduce decoder modality dropout where the model is trained bot"},"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":"2502.01547","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2025-02-03T17:29:52Z","cross_cats_sorted":["cs.CV","cs.SD"],"title_canon_sha256":"fbf9058ef1ba2fa8d5fdc539f7c79d19858ccf12eec80e58feed8092521e563a","abstract_canon_sha256":"acf3c084c660d24f988ca6e23e040a95b067562c55f81ea84048881280a350c9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:59:49.026290Z","signature_b64":"Sd5u9PlqluAxhQERJmgpzbrObess2ZPn5vEDOWfxECw00xlT4HuJQc0TPNIxhzHfHpg27EHdiEUIVZc/G6eKAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"00b70c627d162358189d18da23497640a1576d3d26111c59e2037d2e627df94e","last_reissued_at":"2026-07-05T10:59:49.025824Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:59:49.025824Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"mWhisper-Flamingo for Multilingual Audio-Visual Noise-Robust Speech Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.SD"],"primary_cat":"eess.AS","authors_text":"Andrew Rouditchenko, Hilde Kuehne, James Glass, Rogerio Feris, Samuel Thomas","submitted_at":"2025-02-03T17:29:52Z","abstract_excerpt":"Audio-Visual Speech Recognition (AVSR) combines lip-based video with audio and can improve performance in noise, but most methods are trained only on English data. One limitation is the lack of large-scale multilingual video data, which makes it hard to train models from scratch. In this work, we propose mWhisper-Flamingo for multilingual AVSR which combines the strengths of a pre-trained audio model (Whisper) and video model (AV-HuBERT). To enable better multi-modal integration and improve the noisy multilingual performance, we introduce decoder modality dropout where the model is trained bot"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.01547","kind":"arxiv","version":3},"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/2502.01547/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":"2502.01547","created_at":"2026-07-05T10:59:49.025881+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.01547v3","created_at":"2026-07-05T10:59:49.025881+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.01547","created_at":"2026-07-05T10:59:49.025881+00:00"},{"alias_kind":"pith_short_12","alias_value":"AC3QYYT5CYRV","created_at":"2026-07-05T10:59:49.025881+00:00"},{"alias_kind":"pith_short_16","alias_value":"AC3QYYT5CYRVQGE5","created_at":"2026-07-05T10:59:49.025881+00:00"},{"alias_kind":"pith_short_8","alias_value":"AC3QYYT5","created_at":"2026-07-05T10:59:49.025881+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.03186","citing_title":"CoGenAV: Versatile Audio-Visual Representation Learning via Contrastive-Generative Synchronization","ref_index":7,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AC3QYYT5CYRVQGE5DDNCGSLWIC","json":"https://pith.science/pith/AC3QYYT5CYRVQGE5DDNCGSLWIC.json","graph_json":"https://pith.science/api/pith-number/AC3QYYT5CYRVQGE5DDNCGSLWIC/graph.json","events_json":"https://pith.science/api/pith-number/AC3QYYT5CYRVQGE5DDNCGSLWIC/events.json","paper":"https://pith.science/paper/AC3QYYT5"},"agent_actions":{"view_html":"https://pith.science/pith/AC3QYYT5CYRVQGE5DDNCGSLWIC","download_json":"https://pith.science/pith/AC3QYYT5CYRVQGE5DDNCGSLWIC.json","view_paper":"https://pith.science/paper/AC3QYYT5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.01547&json=true","fetch_graph":"https://pith.science/api/pith-number/AC3QYYT5CYRVQGE5DDNCGSLWIC/graph.json","fetch_events":"https://pith.science/api/pith-number/AC3QYYT5CYRVQGE5DDNCGSLWIC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AC3QYYT5CYRVQGE5DDNCGSLWIC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AC3QYYT5CYRVQGE5DDNCGSLWIC/action/storage_attestation","attest_author":"https://pith.science/pith/AC3QYYT5CYRVQGE5DDNCGSLWIC/action/author_attestation","sign_citation":"https://pith.science/pith/AC3QYYT5CYRVQGE5DDNCGSLWIC/action/citation_signature","submit_replication":"https://pith.science/pith/AC3QYYT5CYRVQGE5DDNCGSLWIC/action/replication_record"}},"created_at":"2026-07-05T10:59:49.025881+00:00","updated_at":"2026-07-05T10:59:49.025881+00:00"}