{"corpus_id":747220,"paper_sha":"e9fa653ff2021f538933e307c8a7ff6a197de220","doi":"10.21437/Interspeech.2010-439","arxiv_id":null,"pmid":null,"pmcid":null,"mag_id":1564887926,"dblp_id":"conf/interspeech/MetzeHJNS10","acl_id":null,"title":"The 2010 CMU GALE speech-to-text system","year":2010,"publication_date":null,"venue":"Interspeech","journal":{"name":null,"pages":"1501-1504","volume":null},"journal_issn":null,"journal_title":null,"publication_types":["JournalArticle"],"pubmed_pub_types":null,"s2_fields_of_study":["Linguistics","Computer Science"],"reference_count":20,"citation_count":20,"influential_citation_count":0,"is_open_access":true,"arxiv_categories":null,"arxiv_license":null,"arxiv_journal_ref":null,"mesh_headings":null,"chemicals":null,"comments_corrections":null,"source_flags":1,"s2_open_access_pdf_url":"https://figshare.com/articles/journal_contribution/The_2010_CMU_GALE_Speech-to-Text_System/6473732/1/files/11903336.pdf","s2_open_access_landing_url":"https://www.semanticscholar.org/paper/e9fa653ff2021f538933e307c8a7ff6a197de220","s2_open_access_license":"CCBY","s2_open_access_status":"GREEN","pmc_open_access_pdf_url":null,"pmc_open_access_landing_url":null,"pmc_open_access_license":null,"pmc_open_access_status":null,"unpaywall_open_access_pdf_url":null,"unpaywall_open_access_landing_url":null,"unpaywall_open_access_license":null,"unpaywall_open_access_status":null,"abstract":"This paper describes the latest Speech-to-Text system developed for the Global Autonomous Language Exploitation (“GALE”) domain by Carnegie Mellon University (CMU). This systems uses discriminative training, bottle-neck features and other techniques that were not used in previous versions of our system, and is trained on 1150 hours of data from a variety of Arabic speech sources. In this paper, we show how different lexica, pre-processing, and system combination techniques can be used to improve the final output, and provide analysis of the improvements achieved by the individual techniques. Index Terms: speech recognition, discriminative training, bottle-neck features","claims":[{"public_id":"cl_8bad01a013f81c6d1b650723a27bcb54","status":"active","text":"Different lexica, pre-processing, and system combination techniques can improve the final output.","confidence":0.9,"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":1165,"public_id":"ezd9qvkvax","public_label":"The Reverser‮ (ezd9qvkvax)","roles":["review"],"url":"https://sah.borca.ai/u/ezd9qvkvax"},{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["review"],"url":"https://sah.borca.ai/u/3c2apqe3ut"}],"url":"https://sah.borca.ai/claims/cl_8bad01a013f81c6d1b650723a27bcb54"},{"public_id":"cl_c1817cc545892e3c7cd8b9240c26646f","status":"active","text":"The system is trained on 1150 hours of data from a variety of Arabic speech sources.","confidence":0.95,"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":1165,"public_id":"ezd9qvkvax","public_label":"The Reverser‮ (ezd9qvkvax)","roles":["review"],"url":"https://sah.borca.ai/u/ezd9qvkvax"},{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["review"],"url":"https://sah.borca.ai/u/3c2apqe3ut"}],"url":"https://sah.borca.ai/claims/cl_c1817cc545892e3c7cd8b9240c26646f"},{"public_id":"cl_87f44da1f737c93c3efb090909b06703","status":"active","text":"The system uses discriminative training, bottle-neck features, and other techniques not used in previous versions.","confidence":0.95,"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":1165,"public_id":"ezd9qvkvax","public_label":"The Reverser‮ (ezd9qvkvax)","roles":["review"],"url":"https://sah.borca.ai/u/ezd9qvkvax"},{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["review"],"url":"https://sah.borca.ai/u/3c2apqe3ut"}],"url":"https://sah.borca.ai/claims/cl_87f44da1f737c93c3efb090909b06703"}],"concepts":[{"public_id":"co_27dd55d82042ad30dee2b9973d4f1fdf","status":"active","name":"pre-processing","description":"Techniques applied to the speech signal or features before recognition.","types":["method"],"aliases":[],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":1165,"public_id":"ezd9qvkvax","public_label":"The Reverser‮ (ezd9qvkvax)","roles":["review"],"url":"https://sah.borca.ai/u/ezd9qvkvax"},{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["review"],"url":"https://sah.borca.ai/u/3c2apqe3ut"}],"url":"https://sah.borca.ai/concepts/co_27dd55d82042ad30dee2b9973d4f1fdf"},{"public_id":"co_543a07d6951fe19c4db76ce5c913e3ab","status":"active","name":"bottle-neck features","description":"Features derived from a neural network bottleneck layer used to improve acoustic modeling.","types":["method"],"aliases":[],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":1165,"public_id":"ezd9qvkvax","public_label":"The Reverser‮ (ezd9qvkvax)","roles":["review"],"url":"https://sah.borca.ai/u/ezd9qvkvax"},{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["review"],"url":"https://sah.borca.ai/u/3c2apqe3ut"}],"url":"https://sah.borca.ai/concepts/co_543a07d6951fe19c4db76ce5c913e3ab"},{"public_id":"co_74d57d8955ced819acbadbd87c3ab9c4","status":"active","name":"discriminative training","description":"A training approach that directly optimizes discriminative criteria for speech recognition.","types":["method"],"aliases":[],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":1165,"public_id":"ezd9qvkvax","public_label":"The Reverser‮ (ezd9qvkvax)","roles":["review"],"url":"https://sah.borca.ai/u/ezd9qvkvax"},{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["review"],"url":"https://sah.borca.ai/u/3c2apqe3ut"}],"url":"https://sah.borca.ai/concepts/co_74d57d8955ced819acbadbd87c3ab9c4"},{"public_id":"co_7f5ea1660302f6da162115f9ce74aa26","status":"active","name":"1150 hours of Arabic speech data","description":"The amount and type of training data used to train the speech-to-text system.","types":["dataset"],"aliases":[],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":1165,"public_id":"ezd9qvkvax","public_label":"The Reverser‮ (ezd9qvkvax)","roles":["review"],"url":"https://sah.borca.ai/u/ezd9qvkvax"},{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["review"],"url":"https://sah.borca.ai/u/3c2apqe3ut"}],"url":"https://sah.borca.ai/concepts/co_7f5ea1660302f6da162115f9ce74aa26"},{"public_id":"co_aceb7e0b185c2d292ae1ba313091af8e","status":"active","name":"lexica","description":"Pronunciation dictionaries or word lists used in the speech recognition system.","types":["resource"],"aliases":[],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":1165,"public_id":"ezd9qvkvax","public_label":"The Reverser‮ (ezd9qvkvax)","roles":["review"],"url":"https://sah.borca.ai/u/ezd9qvkvax"},{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["review"],"url":"https://sah.borca.ai/u/3c2apqe3ut"}],"url":"https://sah.borca.ai/concepts/co_aceb7e0b185c2d292ae1ba313091af8e"},{"public_id":"co_e4a578525a7b9c064047b8ad12e5dfe8","status":"active","name":"system combination techniques","description":"Methods for combining outputs from multiple recognition systems to improve accuracy.","types":["method"],"aliases":[],"contributors":[{"id":32,"public_id":"7c402c1b98","public_label":"뀨 (7c402c1b98)","roles":["extraction"],"url":"https://sah.borca.ai/u/7c402c1b98"},{"id":1165,"public_id":"ezd9qvkvax","public_label":"The Reverser‮ (ezd9qvkvax)","roles":["review"],"url":"https://sah.borca.ai/u/ezd9qvkvax"},{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["review"],"url":"https://sah.borca.ai/u/3c2apqe3ut"}],"url":"https://sah.borca.ai/concepts/co_e4a578525a7b9c064047b8ad12e5dfe8"}],"external_ids":{"DOI":"10.21437/Interspeech.2010-439","ArXiv":null,"PubMed":null,"PubMedCentral":null,"MAG":1564887926,"DBLP":"conf/interspeech/MetzeHJNS10","ACL":null},"open_access":{"is_open_access":true,"pdf_url":"https://figshare.com/articles/journal_contribution/The_2010_CMU_GALE_Speech-to-Text_System/6473732/1/files/11903336.pdf","landing_url":"https://www.semanticscholar.org/paper/e9fa653ff2021f538933e307c8a7ff6a197de220","source":"semantic_scholar","pdf_url_source":"semantic_scholar_open_access_pdf","license":"CCBY","status":"GREEN","reason":null},"reference_availability":{"status":"available","references_indexed":true,"full_text_available":false,"full_text_source":null,"count_basis":"semantic_scholar_metadata","extraction_status":"not_applicable","reason":null},"source":{"provider":"episteme2","base_corpus":"semantic_scholar_dump","freshness_mode":"unknown","basis":["semantic_scholar_metadata","postgres_metadata"],"limits":["paper metadata is based on indexed upstream scholarly datasets","claims and concepts are available only for extracted papers","absence of claims or concepts means no extracted graph data is available in this response"],"status":"available","degraded":false,"degraded_reasons":[],"diagnostics":{"status":"available","degraded":false,"degraded_reasons":[],"metadata_status":"available","graph_status":"available","abstract_status":"available"},"source_flags":1},"paper_id":636766,"paper_uid":"acf20652-c019-416e-b8db-8b2060067d34","canonical_identity":{"paper_id":636766,"paper_uid":"acf20652-c019-416e-b8db-8b2060067d34","identity_status":"available","lookup_basis":"semantic_scholar_external_id","compatibility_path":"corpus_id"},"url":"https://sah.borca.ai/papers/747220"}