{"id":168260,"date":"2015-04-01T00:00:00","date_gmt":"2015-04-01T00:00:00","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/msr-research-item\/an-analysis-of-convolutional-neural-networks-for-speech-recognition\/"},"modified":"2018-10-16T20:12:07","modified_gmt":"2018-10-17T03:12:07","slug":"an-analysis-of-convolutional-neural-networks-for-speech-recognition","status":"publish","type":"msr-research-item","link":"https:\/\/www.microsoft.com\/en-us\/research\/publication\/an-analysis-of-convolutional-neural-networks-for-speech-recognition\/","title":{"rendered":"An Analysis of Convolutional Neural Networks for Speech Recognition"},"content":{"rendered":"<div class=\"asset-content\">\n<p>Despite the fact that several sites have reported the effectiveness of convolutional neural networks (CNNs) on some tasks, there is no deep analysis regarding why CNNs perform well and in which case we should see CNNs \u2019 advantage. In the light of this, this paper aims to provide some detailed analysis of CNNs. By visualizing the localized filters learned in the convolutional layer, we show that edge detectors in varying directions can be automatically learned. We then identify four domains we think CNNs can consistently provide advantages over fully-connected deep neural networks (DNNs): channel-mismatched training-test conditions, noise robustness, distant speech recognition, and low-footprint models. For distant speech recognition, a CNN trained on 1000 hours of Kinect distant speech data obtains relative 4% word error rate reduction (WERR) over a DNN of a similar size. To our knowledge, this is the largest corpus so far reported in the literature for CNNs to show its effectiveness. Lastly, we establish that the CNN structure combined with maxout units is the most effective model under small-sizing constraints for the purpose of deploying small-footprint models to devices. This setup gives relative 9.3% WERR from DNNs with sigmoid units.<\/p>\n<\/div>\n<p><!-- .asset-content --><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Despite the fact that several sites have reported the effectiveness of convolutional neural networks (CNNs) on some tasks, there is no deep analysis regarding why CNNs perform well and in which case we should see CNNs \u2019 advantage. In the light of this, this paper aims to provide some detailed analysis of CNNs. By visualizing [&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":"","footnotes":""},"msr-content-type":[3],"msr-research-highlight":[],"research-area":[13545],"msr-publication-type":[193716],"msr-product-type":[],"msr-focus-area":[],"msr-platform":[],"msr-download-source":[],"msr-locale":[268875],"msr-post-option":[],"msr-field-of-study":[],"msr-conference":[],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-168260","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-human-language-technologies","msr-locale-en_us"],"msr_publishername":"","msr_edition":"ICASSP","msr_affiliation":"","msr_published_date":"2015-04-01","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"","msr_chapter":"","msr_isbn":"","msr_journal":"","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":"217195","msr_publicationurl":"","msr_doi":"","msr_publication_uploader":[{"type":"file","title":"Detailed Analysis of Convolutional Neural Networks for Speech Recognitio_Final.pdf","viewUrl":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2015\/04\/Detailed-Analysis-of-Convolutional-Neural-Networks-for-Speech-Recognitio_Final.pdf","id":217195,"label_id":0}],"msr_related_uploader":"","msr_attachments":[{"id":217195,"url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2015\/04\/Detailed-Analysis-of-Convolutional-Neural-Networks-for-Speech-Recognitio_Final.pdf"}],"msr-author-ordering":[{"type":"text","value":"Jui-Ting Huang","user_id":0,"rest_url":false},{"type":"user_nicename","value":"jinyli","user_id":32312,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=jinyli"},{"type":"user_nicename","value":"ygong","user_id":34994,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=ygong"}],"msr_impact_theme":[],"msr_research_lab":[],"msr_event":[],"msr_group":[],"msr_project":[],"publication":[],"video":[],"download":[],"msr_publication_type":"inproceedings","related_content":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/168260"}],"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\/168260\/revisions"}],"predecessor-version":[{"id":524040,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/168260\/revisions\/524040"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=168260"}],"wp:term":[{"taxonomy":"msr-content-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-content-type?post=168260"},{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=168260"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=168260"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=168260"},{"taxonomy":"msr-product-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-product-type?post=168260"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=168260"},{"taxonomy":"msr-platform","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-platform?post=168260"},{"taxonomy":"msr-download-source","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-download-source?post=168260"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=168260"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=168260"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=168260"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=168260"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=168260"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=168260"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=168260"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}