{"id":157804,"date":"2009-01-01T00:00:00","date_gmt":"2009-01-01T00:00:00","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/msr-research-item\/active-learning-and-semi-supervised-learning-for-speech-recognition-a-unified-framework-using-the-global-entropy-reduction-maximization-criterion\/"},"modified":"2018-10-16T20:56:41","modified_gmt":"2018-10-17T03:56:41","slug":"active-learning-and-semi-supervised-learning-for-speech-recognition-a-unified-framework-using-the-global-entropy-reduction-maximization-criterion","status":"publish","type":"msr-research-item","link":"https:\/\/www.microsoft.com\/en-us\/research\/publication\/active-learning-and-semi-supervised-learning-for-speech-recognition-a-unified-framework-using-the-global-entropy-reduction-maximization-criterion\/","title":{"rendered":"Active Learning and Semi-supervised Learning for Speech Recognition: A Unified Framework using the Global Entropy Reduction Maximization Criterion"},"content":{"rendered":"<div class=\"asset-content\">\n<p>We propose a unified global entropy reduction maximization (GERM) framework for active learning and semi-supervised learning for speech recognition. Active learning aims to select a limited subset of utterances for transcribing from a large amount of un-transcribed utterances, while semi-supervised learning addresses the problem of selecting right transcriptions for un-transcribed utterances, so that the accuracy of the automatic speech recognition system can be maximized. We show that both the traditional confidence-based active learning and semi-supervised learning approaches can be improved by maximizing the lattice entropy reduction over the whole dataset. We introduce our criterion and framework, show how the criterion can be simplified and approximated, and describe how these approaches can be combined. We demonstrate the effectiveness of our new framework and algorithm with directory assistance data collected under the real usage scenarios and show that our GERM based active learning and semi-supervised learning algorithms consistently outperform the confidence-based counterparts by a significant margin. Using our new active learning algorithm cuts the number of utterances needed for transcribing by 50% to achieve the same recognition accuracy obtained using the confidence- based active learning approach, and by 60% compared to the random sampling approach. Using our new semisupervised algorithm we can determine the cutoff point in determining which utterance-transcription pair to use in a principled way by demonstrating that the point it finds is very close to the achievable peak point.<\/p>\n<\/div>\n<p><!-- .asset-content --><\/p>\n","protected":false},"excerpt":{"rendered":"<p>We propose a unified global entropy reduction maximization (GERM) framework for active learning and semi-supervised learning for speech recognition. Active learning aims to select a limited subset of utterances for transcribing from a large amount of un-transcribed utterances, while semi-supervised learning addresses the problem of selecting right transcriptions for un-transcribed utterances, so that the accuracy [&hellip;]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr-author-ordering":null,"msr_publishername":"Elsevier","msr_publisher_other":"","msr_booktitle":"","msr_chapter":"","msr_edition":"","msr_editors":"","msr_how_published":"","msr_isbn":"","msr_issue":"","msr_journal":"Computer Speech and Language - Special Issue on Emergent Artificial Intelligence Approaches for Pattern Recognition in Speech and Language Processing","msr_number":"","msr_organization":"","msr_pages_string":"","msr_page_range_start":"","msr_page_range_end":"","msr_series":"","msr_volume":"","msr_copyright":"","msr_conference_name":"","msr_doi":"","msr_arxiv_id":"","msr_s2_paper_id":"","msr_mag_id":"","msr_pubmed_id":"","msr_other_authors":"Balakrishnan Varadarajan","msr_other_contributors":"","msr_speaker":"","msr_award":"","msr_affiliation":"","msr_institution":"","msr_host":"","msr_version":"","msr_duration":"","msr_original_fields_of_study":"","msr_release_tracker_id":"","msr_s2_match_type":"","msr_citation_count_updated":"","msr_published_date":"2009-01-01","msr_highlight_text":"","msr_notes":"","msr_longbiography":"","msr_publicationurl":"","msr_external_url":"","msr_secondary_video_url":"","msr_conference_url":"","msr_journal_url":"","msr_s2_pdf_url":"","msr_year":2009,"msr_citation_count":0,"msr_influential_citations":0,"msr_reference_count":0,"msr_s2_match_confidence":0,"msr_microsoftintellectualproperty":true,"msr_s2_open_access":false,"msr_s2_author_ids":[],"msr_pub_ids":[],"msr_hide_image_in_river":0,"footnotes":""},"msr-research-highlight":[],"research-area":[13556,13554,13546],"msr-publication-type":[193715],"msr-publisher":[],"msr-focus-area":[],"msr-locale":[268875],"msr-post-option":[],"msr-field-of-study":[],"msr-conference":[],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-157804","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-artificial-intelligence","msr-research-area-human-computer-interaction","msr-research-area-computational-sciences-mathematics","msr-locale-en_us"],"msr_publishername":"Elsevier","msr_edition":"","msr_affiliation":"","msr_published_date":"2009-01-01","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"","msr_chapter":"","msr_isbn":"","msr_journal":"Computer Speech and Language - Special Issue on Emergent Artificial Intelligence Approaches for Pattern Recognition in Speech and Language Processing","msr_volume":"","msr_number":"","msr_editors":"","msr_series":"","msr_issue":"","msr_organization":"","msr_how_published":"","msr_notes":"","msr_highlight_text":"","msr_release_tracker_id":"","msr_original_fields_of_study":"","msr_download_urls":"","msr_external_url":"","msr_secondary_video_url":"","msr_longbiography":"","msr_microsoftintellectualproperty":1,"msr_main_download":"207837","msr_publicationurl":"","msr_doi":"","msr_publication_uploader":[{"type":"file","title":"ActiveSemi-JCSL2009-inpress.pdf","viewUrl":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2016\/02\/ActiveSemi-JCSL2009-inpress.pdf","id":207837,"label_id":0}],"msr_related_uploader":"","msr_citation_count":0,"msr_citation_count_updated":"","msr_s2_paper_id":"","msr_influential_citations":0,"msr_reference_count":0,"msr_arxiv_id":"","msr_s2_author_ids":[],"msr_s2_open_access":false,"msr_s2_pdf_url":null,"msr_attachments":[{"id":207837,"url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2016\/02\/ActiveSemi-JCSL2009-inpress.pdf"}],"msr-author-ordering":[{"type":"user_nicename","value":"dongyu","user_id":31667,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=dongyu"},{"type":"text","value":"Balakrishnan Varadarajan","user_id":0,"rest_url":false},{"type":"user_nicename","value":"deng","user_id":31602,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=deng"},{"type":"user_nicename","value":"alexac","user_id":30932,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=alexac"}],"msr_impact_theme":[],"msr_research_lab":[],"msr_event":[],"msr_group":[],"msr_project":[170140,169434],"publication":[],"video":[],"msr-tool":[],"msr_publication_type":"article","related_content":{"projects":[{"ID":170140,"post_title":"Voice Search: Say What You Want and Get It","post_name":"voice-search-say-what-you-want-and-get-it","post_type":"msr-project","post_date":"2008-12-15 13:28:48","post_modified":"2019-08-19 15:35:11","post_status":"publish","permalink":"https:\/\/www.microsoft.com\/en-us\/research\/project\/voice-search-say-what-you-want-and-get-it\/","post_excerpt":"In the Voice Search project, we envision a future where you can ask your cellphone for any kind of information and get it. With a small cellphone, there is a heavy tax on traditional keyboard based information entry, and we believe it can be significantly more convenient to communicate by voice. Our work focuses on making this communication more reliable, and able to cover the full range of information needed in daily life.","_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/170140"}]}},{"ID":169434,"post_title":"Acoustic Modeling","post_name":"acoustic-modeling","post_type":"msr-project","post_date":"2004-01-29 16:42:42","post_modified":"2019-08-14 14:50:04","post_status":"publish","permalink":"https:\/\/www.microsoft.com\/en-us\/research\/project\/acoustic-modeling\/","post_excerpt":"Acoustic modeling of speech typically refers to the process of\u00a0establishing statistical\u00a0representations for the feature vector sequences\u00a0computed from the speech waveform. Hidden Markov Model (HMM) is one most common type of acoustuc models. Other acosutic models include segmental models, super-segmental models (including hidden dynamic models), neural networks, maximum entropy models, and (hidden) conditional random fields, etc. Acoustic modeling also encompasses \"pronunciation modeling\", which describes how a sequence or multi-sequences of fundamental speech units\u00a0(such as phones or&hellip;","_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/169434"}]}}]},"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/157804","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-research-item"}],"version-history":[{"count":1,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/157804\/revisions"}],"predecessor-version":[{"id":435054,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/157804\/revisions\/435054"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=157804"}],"wp:term":[{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=157804"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=157804"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=157804"},{"taxonomy":"msr-publisher","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publisher?post=157804"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=157804"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=157804"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=157804"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=157804"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=157804"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=157804"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=157804"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=157804"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}