{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:ATQNMF3BRTHFSNLMWR5RCPGA66","short_pith_number":"pith:ATQNMF3B","schema_version":"1.0","canonical_sha256":"04e0d617618cce59356cb47b113cc0f7a79cabc7eab05b294017f1849e1806a6","source":{"kind":"arxiv","id":"2406.10082","version":3},"attestation_state":"computed","paper":{"title":"Whisper-Flamingo: Integrating Visual Features into Whisper for Audio-Visual Speech Recognition and Translation","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, Leonid Karlinsky, Rogerio Feris, Samuel Thomas, Yuan Gong","submitted_at":"2024-06-14T14:36:54Z","abstract_excerpt":"Audio-Visual Speech Recognition (AVSR) uses lip-based video to improve performance in noise. Since videos are harder to obtain than audio, the video training data of AVSR models is usually limited to a few thousand hours. In contrast, speech models such as Whisper are trained with hundreds of thousands of hours of data, and thus learn a better speech-to-text decoder. The huge training data difference motivates us to adapt Whisper to handle video inputs. Inspired by Flamingo which injects visual features into language models, we propose Whisper-Flamingo which integrates visual features into the"},"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":"2406.10082","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2024-06-14T14:36:54Z","cross_cats_sorted":["cs.CV","cs.SD"],"title_canon_sha256":"1830c94b2deb42293954c80fbfb501b3d41fc3a22100b95bc2ac604e62721ccf","abstract_canon_sha256":"57825a3d40148e81807a62b4601c047ded19442ed3161794dde15d5467ebfb59"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:37:54.987018Z","signature_b64":"44j6dAF/jMR1zYojJzymm6VzmFQ54+260FibvtLxa0OF3E5oR6lHz0qP5LY17DLNSDRTVXgPqWKAompP/4XjAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"04e0d617618cce59356cb47b113cc0f7a79cabc7eab05b294017f1849e1806a6","last_reissued_at":"2026-07-05T09:37:54.986503Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:37:54.986503Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Whisper-Flamingo: Integrating Visual Features into Whisper for Audio-Visual Speech Recognition and Translation","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, Leonid Karlinsky, Rogerio Feris, Samuel Thomas, Yuan Gong","submitted_at":"2024-06-14T14:36:54Z","abstract_excerpt":"Audio-Visual Speech Recognition (AVSR) uses lip-based video to improve performance in noise. Since videos are harder to obtain than audio, the video training data of AVSR models is usually limited to a few thousand hours. In contrast, speech models such as Whisper are trained with hundreds of thousands of hours of data, and thus learn a better speech-to-text decoder. The huge training data difference motivates us to adapt Whisper to handle video inputs. Inspired by Flamingo which injects visual features into language models, we propose Whisper-Flamingo which integrates visual features into the"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.10082","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/2406.10082/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":"2406.10082","created_at":"2026-07-05T09:37:54.986566+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.10082v3","created_at":"2026-07-05T09:37:54.986566+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.10082","created_at":"2026-07-05T09:37:54.986566+00:00"},{"alias_kind":"pith_short_12","alias_value":"ATQNMF3BRTHF","created_at":"2026-07-05T09:37:54.986566+00:00"},{"alias_kind":"pith_short_16","alias_value":"ATQNMF3BRTHFSNLM","created_at":"2026-07-05T09:37:54.986566+00:00"},{"alias_kind":"pith_short_8","alias_value":"ATQNMF3B","created_at":"2026-07-05T09:37:54.986566+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.07294","citing_title":"Multimodal Voice Activity Projection for Turn-Taking in Social Robots with Voice-Activity-Related Pretrained Encoders","ref_index":31,"is_internal_anchor":true},{"citing_arxiv_id":"2603.21664","citing_title":"HumanOmni-Speaker: Identifying Who said What and When","ref_index":36,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ATQNMF3BRTHFSNLMWR5RCPGA66","json":"https://pith.science/pith/ATQNMF3BRTHFSNLMWR5RCPGA66.json","graph_json":"https://pith.science/api/pith-number/ATQNMF3BRTHFSNLMWR5RCPGA66/graph.json","events_json":"https://pith.science/api/pith-number/ATQNMF3BRTHFSNLMWR5RCPGA66/events.json","paper":"https://pith.science/paper/ATQNMF3B"},"agent_actions":{"view_html":"https://pith.science/pith/ATQNMF3BRTHFSNLMWR5RCPGA66","download_json":"https://pith.science/pith/ATQNMF3BRTHFSNLMWR5RCPGA66.json","view_paper":"https://pith.science/paper/ATQNMF3B","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.10082&json=true","fetch_graph":"https://pith.science/api/pith-number/ATQNMF3BRTHFSNLMWR5RCPGA66/graph.json","fetch_events":"https://pith.science/api/pith-number/ATQNMF3BRTHFSNLMWR5RCPGA66/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ATQNMF3BRTHFSNLMWR5RCPGA66/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ATQNMF3BRTHFSNLMWR5RCPGA66/action/storage_attestation","attest_author":"https://pith.science/pith/ATQNMF3BRTHFSNLMWR5RCPGA66/action/author_attestation","sign_citation":"https://pith.science/pith/ATQNMF3BRTHFSNLMWR5RCPGA66/action/citation_signature","submit_replication":"https://pith.science/pith/ATQNMF3BRTHFSNLMWR5RCPGA66/action/replication_record"}},"created_at":"2026-07-05T09:37:54.986566+00:00","updated_at":"2026-07-05T09:37:54.986566+00:00"}