{"id":81,"date":"2018-11-23T10:48:27","date_gmt":"2018-11-23T10:48:27","guid":{"rendered":"http:\/\/speechlabs.pl\/?page_id=81"},"modified":"2018-12-07T20:21:34","modified_gmt":"2018-12-07T20:21:34","slug":"arm","status":"publish","type":"page","link":"https:\/\/speechlabs.pl\/en\/our-offer\/arm\/","title":{"rendered":"ARM"},"content":{"rendered":"<p>\r\n\t\t<b>ARM is an automatic speech recognition engine for continuous speech in Polish. It is based on Hidden Markov Models (HMM) and dedicated  statistical acoustic and linguistic models. <\/b>\r\n\t<\/p>\r\n\t<img decoding=\"async\" class=\"engine\" src=\"http:\/\/speechlabs.pl\/wp-content\/uploads\/2018\/11\/about_en.png\">\r\n\t<label class=\"img-label\">ARM engine scheme<\/label>\r\n\t<p>\r\n\t  <b>Speech recognition results<\/b> contain:\r\n\t  <ul class=\"circle\">\r\n\t    <li>transcript in the form of a sequence of words,<\/li>\r\n\t    <li>paritioning into segments,<\/li>\r\n\t    <li>time stamps marking beginning and end of each word,<\/li>\r\n\t    <li>confidence measure,<\/li>\r\n\t    <li>set of alternative hypotheses.<\/li>\r\n\t  <\/ul>\r\n<\/p>\r\n\t<p>\r\n\t  ARM engine is used in <a href=\"http:\/\/speechlabs.pl\/oferta\/airnote\/\">Airnote<\/a> and <a href=\"http:\/\/speechlabs.pl\/oferta\/tin\/\">TIN<\/a>. It can also be integrated with users&#8217; systems as a speech recognition service. Such service is implemented in:\r\n\t  <ul class=\"circle\">\r\n\t   <li>Content Analysis System for Television <a href=\"http:\/\/cast.info.pl\/\" target=\"_blank\">CAST<\/a> deployed at Department of Journalism and Political Sciences, Adam Mickiewicz University, <\/li>\r\n\t   <li>Electronic Media Monitoring and Analysis system supporting open source intelligence, <\/li>\r\n\t   <li><a href=\"http:\/\/psmm.pl\/pl\" target=\"_blank\">Press Service Media Monitoring<\/a> system,<\/li>\r\n\t    <li>IVA service platform of virtual voice agents for automation of emergency hotline centers.<\/li>\r\n\t  <\/ul> \r\n<\/p>\r\n\t  <p>\r\n            The ARM engine can be <b>adjusted to a specific application area<\/b> through acoustic and linguistic model adaptation based on recordings and text corpora provided by users. User can also define rules for speech recognition transformation to a desired presentation form.    \r\n\t  <\/p><br\/>\r\n\t  <p>\r\nSpeech recognition service is deployed in users&#8217; internal network and does not require communication with any external services. Hence <b>data confidentiality<\/b> can be ensured.\r\n\t \r\n\t<\/p><br\/>\r\n\t<p>\r\n\t\tARM engine is a result of <a href=\"http:\/\/speechlabs.pl\/en\/completed\/#zskm\">ZSKM<\/a> project. It was further developed in <a href=\"http:\/\/speechlabs.pl\/en\/completed\/#arm1\">ARM-1<\/a>. It is a Windows based solution. \r\n\t<\/p>","protected":false},"excerpt":{"rendered":"ARM is an automatic speech recognition engine for continuous speech in Polish. It is based on Hidden Markov Models (HMM) and dedicated statistical acoustic and linguistic models. ARM engine scheme Speech recognition results contain: transcript in the form of a sequence of words, paritioning into segments, time stamps marking beginning and end of each word, [&hellip;]","protected":false},"author":1,"featured_media":0,"parent":70,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_locale":"en_US","_original_post":"56","footnotes":""},"class_list":["post-81","page","type-page","status-publish","hentry","en-US"],"meta_box":{"parallax_image":{"width":150,"height":150,"file":"2018\/12\/IMG_2447.jpg","sizes":{"thumbnail":{"file":"IMG_2447-150x150.jpg","width":150,"height":150,"mime-type":"image\/jpeg","url":"https:\/\/speechlabs.pl\/wp-content\/uploads\/2018\/12\/IMG_2447-150x150.jpg"},"medium":{"file":"IMG_2447-300x200.jpg","width":300,"height":200,"mime-type":"image\/jpeg","url":"https:\/\/speechlabs.pl\/wp-content\/uploads\/2018\/12\/IMG_2447-300x200.jpg"},"medium_large":{"file":"IMG_2447-768x512.jpg","width":768,"height":512,"mime-type":"image\/jpeg","url":"https:\/\/speechlabs.pl\/wp-content\/uploads\/2018\/12\/IMG_2447-768x512.jpg"},"large":{"file":"IMG_2447-1024x682.jpg","width":1024,"height":682,"mime-type":"image\/jpeg","url":"https:\/\/speechlabs.pl\/wp-content\/uploads\/2018\/12\/IMG_2447-1024x682.jpg"}},"image_meta":{"aperture":"3.5","credit":"","camera":"Canon EOS 7D","caption":"","created_timestamp":"1455725920","copyright":"","focal_length":"18","iso":"1250","shutter_speed":"0.01","title":"","orientation":"1","keywords":[]},"ID":"482","name":"IMG_2447.jpg","path":"\/home\/sftpusers\/speechlabs\/www\/wp-content\/uploads\/2018\/12\/IMG_2447.jpg","url":"https:\/\/speechlabs.pl\/wp-content\/uploads\/2018\/12\/IMG_2447-150x150.jpg","full_url":"https:\/\/speechlabs.pl\/wp-content\/uploads\/2018\/12\/IMG_2447.jpg","title":"IMG_2447","caption":"","description":"","alt":"","srcset":false},"parallax_position":"center center","hide_news":"0","extra_content":"","custom_title":"","custom_title_link":"","enable_slider":"0","slide":[{"_state":"collapsed"}]},"_links":{"self":[{"href":"https:\/\/speechlabs.pl\/wp-json\/wp\/v2\/pages\/81","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/speechlabs.pl\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/speechlabs.pl\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/speechlabs.pl\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/speechlabs.pl\/wp-json\/wp\/v2\/comments?post=81"}],"version-history":[{"count":12,"href":"https:\/\/speechlabs.pl\/wp-json\/wp\/v2\/pages\/81\/revisions"}],"predecessor-version":[{"id":686,"href":"https:\/\/speechlabs.pl\/wp-json\/wp\/v2\/pages\/81\/revisions\/686"}],"up":[{"embeddable":true,"href":"https:\/\/speechlabs.pl\/wp-json\/wp\/v2\/pages\/70"}],"wp:attachment":[{"href":"https:\/\/speechlabs.pl\/wp-json\/wp\/v2\/media?parent=81"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}