{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:TKKNP7SOYFCCVRMRJW7524WURE","short_pith_number":"pith:TKKNP7SO","schema_version":"1.0","canonical_sha256":"9a94d7fe4ec1442ac5914dbfdd72d48901d9336e794d46d0ce0fcaf214dc83a9","source":{"kind":"arxiv","id":"2403.04245","version":1},"attestation_state":"computed","paper":{"title":"A Study of Dropout-Induced Modality Bias on Robustness to Missing Video Frames for Audio-Visual Speech Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.LG","cs.MM","eess.AS"],"primary_cat":"cs.SD","authors_text":"Chin-Hui Lee, Hang Chen, Haotian Wang, Jiefeng Ma, Jun Du, Ruoyu Wang, Shihao Chen, Yusheng Dai","submitted_at":"2024-03-07T06:06:55Z","abstract_excerpt":"Advanced Audio-Visual Speech Recognition (AVSR) systems have been observed to be sensitive to missing video frames, performing even worse than single-modality models. While applying the dropout technique to the video modality enhances robustness to missing frames, it simultaneously results in a performance loss when dealing with complete data input. In this paper, we investigate this contrasting phenomenon from the perspective of modality bias and reveal that an excessive modality bias on the audio caused by dropout is the underlying reason. Moreover, we present the Modality Bias Hypothesis (M"},"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":"2403.04245","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SD","submitted_at":"2024-03-07T06:06:55Z","cross_cats_sorted":["cs.CV","cs.LG","cs.MM","eess.AS"],"title_canon_sha256":"4dea1a33b2d40e1c4e5db8bfe210796fdee793d542225c51fa8906aa57a3f25f","abstract_canon_sha256":"c4b410b7d64083881e87d09f11aef43d25ed2dbccf9d7251f87baebb5b26378e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:53:20.524984Z","signature_b64":"MWVsNi4NwNpPkEXcONobqGzJYkABtQPPUgJwkr8fnlzolBSbGZ0H1pZpMV69sJnkCOqBUBk6ucdlIBlID4QhCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9a94d7fe4ec1442ac5914dbfdd72d48901d9336e794d46d0ce0fcaf214dc83a9","last_reissued_at":"2026-07-05T07:53:20.524510Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:53:20.524510Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Study of Dropout-Induced Modality Bias on Robustness to Missing Video Frames for Audio-Visual Speech Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.LG","cs.MM","eess.AS"],"primary_cat":"cs.SD","authors_text":"Chin-Hui Lee, Hang Chen, Haotian Wang, Jiefeng Ma, Jun Du, Ruoyu Wang, Shihao Chen, Yusheng Dai","submitted_at":"2024-03-07T06:06:55Z","abstract_excerpt":"Advanced Audio-Visual Speech Recognition (AVSR) systems have been observed to be sensitive to missing video frames, performing even worse than single-modality models. While applying the dropout technique to the video modality enhances robustness to missing frames, it simultaneously results in a performance loss when dealing with complete data input. In this paper, we investigate this contrasting phenomenon from the perspective of modality bias and reveal that an excessive modality bias on the audio caused by dropout is the underlying reason. Moreover, we present the Modality Bias Hypothesis (M"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.04245","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/2403.04245/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":"2403.04245","created_at":"2026-07-05T07:53:20.524568+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.04245v1","created_at":"2026-07-05T07:53:20.524568+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.04245","created_at":"2026-07-05T07:53:20.524568+00:00"},{"alias_kind":"pith_short_12","alias_value":"TKKNP7SOYFCC","created_at":"2026-07-05T07:53:20.524568+00:00"},{"alias_kind":"pith_short_16","alias_value":"TKKNP7SOYFCCVRMR","created_at":"2026-07-05T07:53:20.524568+00:00"},{"alias_kind":"pith_short_8","alias_value":"TKKNP7SO","created_at":"2026-07-05T07:53:20.524568+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TKKNP7SOYFCCVRMRJW7524WURE","json":"https://pith.science/pith/TKKNP7SOYFCCVRMRJW7524WURE.json","graph_json":"https://pith.science/api/pith-number/TKKNP7SOYFCCVRMRJW7524WURE/graph.json","events_json":"https://pith.science/api/pith-number/TKKNP7SOYFCCVRMRJW7524WURE/events.json","paper":"https://pith.science/paper/TKKNP7SO"},"agent_actions":{"view_html":"https://pith.science/pith/TKKNP7SOYFCCVRMRJW7524WURE","download_json":"https://pith.science/pith/TKKNP7SOYFCCVRMRJW7524WURE.json","view_paper":"https://pith.science/paper/TKKNP7SO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.04245&json=true","fetch_graph":"https://pith.science/api/pith-number/TKKNP7SOYFCCVRMRJW7524WURE/graph.json","fetch_events":"https://pith.science/api/pith-number/TKKNP7SOYFCCVRMRJW7524WURE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TKKNP7SOYFCCVRMRJW7524WURE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TKKNP7SOYFCCVRMRJW7524WURE/action/storage_attestation","attest_author":"https://pith.science/pith/TKKNP7SOYFCCVRMRJW7524WURE/action/author_attestation","sign_citation":"https://pith.science/pith/TKKNP7SOYFCCVRMRJW7524WURE/action/citation_signature","submit_replication":"https://pith.science/pith/TKKNP7SOYFCCVRMRJW7524WURE/action/replication_record"}},"created_at":"2026-07-05T07:53:20.524568+00:00","updated_at":"2026-07-05T07:53:20.524568+00:00"}