{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:VJPZIDAEJVQF7677MFLI4NYILZ","short_pith_number":"pith:VJPZIDAE","schema_version":"1.0","canonical_sha256":"aa5f940c044d605ffbff61568e37085e696a08db6f5d44ff592a5e786a1caf5e","source":{"kind":"arxiv","id":"2105.05920","version":2},"attestation_state":"computed","paper":{"title":"Attention-based Neural Beamforming Layers for Multi-channel Speech Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Bhargav Pulugundla, Brian King, Gokce Keskin, Harish Mallidi, Jasha Droppo, Minhua Wu, Roland Maas, Yang Gao","submitted_at":"2021-05-12T19:32:24Z","abstract_excerpt":"Attention-based beamformers have recently been shown to be effective for multi-channel speech recognition. However, they are less capable at capturing local information. In this work, we propose a 2D Conv-Attention module which combines convolution neural networks with attention for beamforming. We apply self- and cross-attention to explicitly model the correlations within and between the input channels. The end-to-end 2D Conv-Attention model is compared with a multi-head self-attention and superdirective-based neural beamformers. We train and evaluate on an in-house multi-channel dataset. 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":"2105.05920","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2021-05-12T19:32:24Z","cross_cats_sorted":["cs.LG","cs.SD"],"title_canon_sha256":"c225598442a759a311b12422b6ce82fd3d762599852120463930e5e70cd95621","abstract_canon_sha256":"7e5939e5d4942f0ba277eb7cf33a7b8e26e8c2da59230104ea0ed5cb1a9423ed"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:40:35.675003Z","signature_b64":"LDbqzfKZswydLvxEPLsy6BSRqJItq0qKZdabq2pwZ7bgb9fNKLjCDuf/GmO47rY9t2bABHcApKtSLeWeaggDBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aa5f940c044d605ffbff61568e37085e696a08db6f5d44ff592a5e786a1caf5e","last_reissued_at":"2026-07-05T02:40:35.674614Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:40:35.674614Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Attention-based Neural Beamforming Layers for Multi-channel Speech Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Bhargav Pulugundla, Brian King, Gokce Keskin, Harish Mallidi, Jasha Droppo, Minhua Wu, Roland Maas, Yang Gao","submitted_at":"2021-05-12T19:32:24Z","abstract_excerpt":"Attention-based beamformers have recently been shown to be effective for multi-channel speech recognition. However, they are less capable at capturing local information. In this work, we propose a 2D Conv-Attention module which combines convolution neural networks with attention for beamforming. We apply self- and cross-attention to explicitly model the correlations within and between the input channels. The end-to-end 2D Conv-Attention model is compared with a multi-head self-attention and superdirective-based neural beamformers. We train and evaluate on an in-house multi-channel dataset. The"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2105.05920","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/2105.05920/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":"2105.05920","created_at":"2026-07-05T02:40:35.674664+00:00"},{"alias_kind":"arxiv_version","alias_value":"2105.05920v2","created_at":"2026-07-05T02:40:35.674664+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2105.05920","created_at":"2026-07-05T02:40:35.674664+00:00"},{"alias_kind":"pith_short_12","alias_value":"VJPZIDAEJVQF","created_at":"2026-07-05T02:40:35.674664+00:00"},{"alias_kind":"pith_short_16","alias_value":"VJPZIDAEJVQF7677","created_at":"2026-07-05T02:40:35.674664+00:00"},{"alias_kind":"pith_short_8","alias_value":"VJPZIDAE","created_at":"2026-07-05T02:40:35.674664+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/VJPZIDAEJVQF7677MFLI4NYILZ","json":"https://pith.science/pith/VJPZIDAEJVQF7677MFLI4NYILZ.json","graph_json":"https://pith.science/api/pith-number/VJPZIDAEJVQF7677MFLI4NYILZ/graph.json","events_json":"https://pith.science/api/pith-number/VJPZIDAEJVQF7677MFLI4NYILZ/events.json","paper":"https://pith.science/paper/VJPZIDAE"},"agent_actions":{"view_html":"https://pith.science/pith/VJPZIDAEJVQF7677MFLI4NYILZ","download_json":"https://pith.science/pith/VJPZIDAEJVQF7677MFLI4NYILZ.json","view_paper":"https://pith.science/paper/VJPZIDAE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2105.05920&json=true","fetch_graph":"https://pith.science/api/pith-number/VJPZIDAEJVQF7677MFLI4NYILZ/graph.json","fetch_events":"https://pith.science/api/pith-number/VJPZIDAEJVQF7677MFLI4NYILZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VJPZIDAEJVQF7677MFLI4NYILZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VJPZIDAEJVQF7677MFLI4NYILZ/action/storage_attestation","attest_author":"https://pith.science/pith/VJPZIDAEJVQF7677MFLI4NYILZ/action/author_attestation","sign_citation":"https://pith.science/pith/VJPZIDAEJVQF7677MFLI4NYILZ/action/citation_signature","submit_replication":"https://pith.science/pith/VJPZIDAEJVQF7677MFLI4NYILZ/action/replication_record"}},"created_at":"2026-07-05T02:40:35.674664+00:00","updated_at":"2026-07-05T02:40:35.674664+00:00"}