{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:QXRWXOX62WFUAT22PPZKSXXRAU","short_pith_number":"pith:QXRWXOX6","schema_version":"1.0","canonical_sha256":"85e36bbafed58b404f5a7bf2a95ef1051ea5e068f9868255b3bbd147cccafc0e","source":{"kind":"arxiv","id":"2210.17017","version":2},"attestation_state":"computed","paper":{"title":"Blank Collapse: Compressing CTC emission for the faster decoding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Minkyu Jung, Ohhyeok Kwon, Seunghyun Seo, Soonshin Seo","submitted_at":"2022-10-31T02:12:51Z","abstract_excerpt":"Connectionist Temporal Classification (CTC) model is a very efficient method for modeling sequences, especially for speech data. In order to use CTC model as an Automatic Speech Recognition (ASR) task, the beam search decoding with an external language model like n-gram LM is necessary to obtain reasonable results. In this paper we analyze the blank label in CTC beam search deeply and propose a very simple method to reduce the amount of calculation resulting in faster beam search decoding speed. With this method, we can get up to 78% faster decoding speed than ordinary beam search decoding wit"},"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":"2210.17017","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-10-31T02:12:51Z","cross_cats_sorted":["cs.SD","eess.AS"],"title_canon_sha256":"042ef3f26258a0a70c5ac657b57ca31cb47cbbe2137e06cc251ba362fa01812a","abstract_canon_sha256":"789a14842b84497081eee975850a7a28b748f98fe5840bfb02aa00783a1e25ee"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:24:45.279491Z","signature_b64":"k+RVpfINfOQvMsRb9gVsV88lqfygLuKqkLI6VZpCIGWn+YKjq2R9TNl8UdR7JooVXdcMUhtDdUiDTFJvYniNDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"85e36bbafed58b404f5a7bf2a95ef1051ea5e068f9868255b3bbd147cccafc0e","last_reissued_at":"2026-07-05T06:24:45.279063Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:24:45.279063Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Blank Collapse: Compressing CTC emission for the faster decoding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Minkyu Jung, Ohhyeok Kwon, Seunghyun Seo, Soonshin Seo","submitted_at":"2022-10-31T02:12:51Z","abstract_excerpt":"Connectionist Temporal Classification (CTC) model is a very efficient method for modeling sequences, especially for speech data. In order to use CTC model as an Automatic Speech Recognition (ASR) task, the beam search decoding with an external language model like n-gram LM is necessary to obtain reasonable results. In this paper we analyze the blank label in CTC beam search deeply and propose a very simple method to reduce the amount of calculation resulting in faster beam search decoding speed. With this method, we can get up to 78% faster decoding speed than ordinary beam search decoding wit"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.17017","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/2210.17017/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":"2210.17017","created_at":"2026-07-05T06:24:45.279117+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.17017v2","created_at":"2026-07-05T06:24:45.279117+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.17017","created_at":"2026-07-05T06:24:45.279117+00:00"},{"alias_kind":"pith_short_12","alias_value":"QXRWXOX62WFU","created_at":"2026-07-05T06:24:45.279117+00:00"},{"alias_kind":"pith_short_16","alias_value":"QXRWXOX62WFUAT22","created_at":"2026-07-05T06:24:45.279117+00:00"},{"alias_kind":"pith_short_8","alias_value":"QXRWXOX6","created_at":"2026-07-05T06:24:45.279117+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.08384","citing_title":"TASU2: Controllable CTC Simulation for Alignment and Low-Resource Adaptation of Speech LLMs","ref_index":22,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QXRWXOX62WFUAT22PPZKSXXRAU","json":"https://pith.science/pith/QXRWXOX62WFUAT22PPZKSXXRAU.json","graph_json":"https://pith.science/api/pith-number/QXRWXOX62WFUAT22PPZKSXXRAU/graph.json","events_json":"https://pith.science/api/pith-number/QXRWXOX62WFUAT22PPZKSXXRAU/events.json","paper":"https://pith.science/paper/QXRWXOX6"},"agent_actions":{"view_html":"https://pith.science/pith/QXRWXOX62WFUAT22PPZKSXXRAU","download_json":"https://pith.science/pith/QXRWXOX62WFUAT22PPZKSXXRAU.json","view_paper":"https://pith.science/paper/QXRWXOX6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.17017&json=true","fetch_graph":"https://pith.science/api/pith-number/QXRWXOX62WFUAT22PPZKSXXRAU/graph.json","fetch_events":"https://pith.science/api/pith-number/QXRWXOX62WFUAT22PPZKSXXRAU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QXRWXOX62WFUAT22PPZKSXXRAU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QXRWXOX62WFUAT22PPZKSXXRAU/action/storage_attestation","attest_author":"https://pith.science/pith/QXRWXOX62WFUAT22PPZKSXXRAU/action/author_attestation","sign_citation":"https://pith.science/pith/QXRWXOX62WFUAT22PPZKSXXRAU/action/citation_signature","submit_replication":"https://pith.science/pith/QXRWXOX62WFUAT22PPZKSXXRAU/action/replication_record"}},"created_at":"2026-07-05T06:24:45.279117+00:00","updated_at":"2026-07-05T06:24:45.279117+00:00"}