{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2016:HGQZ4LH7Q5ECH7GAYYQT2FSBUR","short_pith_number":"pith:HGQZ4LH7","schema_version":"1.0","canonical_sha256":"39a19e2cff874823fcc0c6213d1641a44a6ecb6f45bc3a8773125e8a786d0fb3","source":{"kind":"arxiv","id":"1610.05688","version":1},"attestation_state":"computed","paper":{"title":"Low-rank and Sparse Soft Targets to Learn Better DNN Acoustic Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.HC","cs.LG"],"primary_cat":"cs.CL","authors_text":"Afsaneh Asaei, Herve Bourlard, Pranay Dighe","submitted_at":"2016-10-18T16:02:10Z","abstract_excerpt":"Conventional deep neural networks (DNN) for speech acoustic modeling rely on Gaussian mixture models (GMM) and hidden Markov model (HMM) to obtain binary class labels as the targets for DNN training. Subword classes in speech recognition systems correspond to context-dependent tied states or senones. The present work addresses some limitations of GMM-HMM senone alignments for DNN training. We hypothesize that the senone probabilities obtained from a DNN trained with binary labels can provide more accurate targets to learn better acoustic models. However, DNN outputs bear inaccuracies which are"},"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":"1610.05688","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2016-10-18T16:02:10Z","cross_cats_sorted":["cs.AI","cs.HC","cs.LG"],"title_canon_sha256":"c643a693e29ce5473f7ed06f29269e52f6a13bb7bd43a29e471f073608b8996b","abstract_canon_sha256":"3eea03f2201b4dc7bd4bb5feb5eac1adccf8a6ae126c00ed7be0d89d1e7d7ab1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:35:55.527865Z","signature_b64":"E/YAu96XHPQOCr/lVzDcBGyRPMYWz6r/ivIcu3Jc5+1U4vr8TcetNzi736ISu11uTI6byBzLMnOmSszTz03EBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"39a19e2cff874823fcc0c6213d1641a44a6ecb6f45bc3a8773125e8a786d0fb3","last_reissued_at":"2026-05-18T00:35:55.527143Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:35:55.527143Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Low-rank and Sparse Soft Targets to Learn Better DNN Acoustic Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.HC","cs.LG"],"primary_cat":"cs.CL","authors_text":"Afsaneh Asaei, Herve Bourlard, Pranay Dighe","submitted_at":"2016-10-18T16:02:10Z","abstract_excerpt":"Conventional deep neural networks (DNN) for speech acoustic modeling rely on Gaussian mixture models (GMM) and hidden Markov model (HMM) to obtain binary class labels as the targets for DNN training. Subword classes in speech recognition systems correspond to context-dependent tied states or senones. The present work addresses some limitations of GMM-HMM senone alignments for DNN training. We hypothesize that the senone probabilities obtained from a DNN trained with binary labels can provide more accurate targets to learn better acoustic models. However, DNN outputs bear inaccuracies which are"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1610.05688","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":""},"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":"1610.05688","created_at":"2026-05-18T00:35:55.527256+00:00"},{"alias_kind":"arxiv_version","alias_value":"1610.05688v1","created_at":"2026-05-18T00:35:55.527256+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1610.05688","created_at":"2026-05-18T00:35:55.527256+00:00"},{"alias_kind":"pith_short_12","alias_value":"HGQZ4LH7Q5EC","created_at":"2026-05-18T12:30:19.053100+00:00"},{"alias_kind":"pith_short_16","alias_value":"HGQZ4LH7Q5ECH7GA","created_at":"2026-05-18T12:30:19.053100+00:00"},{"alias_kind":"pith_short_8","alias_value":"HGQZ4LH7","created_at":"2026-05-18T12:30:19.053100+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/HGQZ4LH7Q5ECH7GAYYQT2FSBUR","json":"https://pith.science/pith/HGQZ4LH7Q5ECH7GAYYQT2FSBUR.json","graph_json":"https://pith.science/api/pith-number/HGQZ4LH7Q5ECH7GAYYQT2FSBUR/graph.json","events_json":"https://pith.science/api/pith-number/HGQZ4LH7Q5ECH7GAYYQT2FSBUR/events.json","paper":"https://pith.science/paper/HGQZ4LH7"},"agent_actions":{"view_html":"https://pith.science/pith/HGQZ4LH7Q5ECH7GAYYQT2FSBUR","download_json":"https://pith.science/pith/HGQZ4LH7Q5ECH7GAYYQT2FSBUR.json","view_paper":"https://pith.science/paper/HGQZ4LH7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1610.05688&json=true","fetch_graph":"https://pith.science/api/pith-number/HGQZ4LH7Q5ECH7GAYYQT2FSBUR/graph.json","fetch_events":"https://pith.science/api/pith-number/HGQZ4LH7Q5ECH7GAYYQT2FSBUR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HGQZ4LH7Q5ECH7GAYYQT2FSBUR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HGQZ4LH7Q5ECH7GAYYQT2FSBUR/action/storage_attestation","attest_author":"https://pith.science/pith/HGQZ4LH7Q5ECH7GAYYQT2FSBUR/action/author_attestation","sign_citation":"https://pith.science/pith/HGQZ4LH7Q5ECH7GAYYQT2FSBUR/action/citation_signature","submit_replication":"https://pith.science/pith/HGQZ4LH7Q5ECH7GAYYQT2FSBUR/action/replication_record"}},"created_at":"2026-05-18T00:35:55.527256+00:00","updated_at":"2026-05-18T00:35:55.527256+00:00"}