<html><body><div style="font-family: arial, helvetica, sans-serif; font-size: 12pt; color: #000000"><div> <!--StartFragment--><div id="post-1954" class="clearfix post post-1954 type-post status-publish format-standard hentry category-job-offers item-wrap" style="box-sizing: border-box; zoom: 1; background: #dddddd; border: none; box-shadow: none; padding: 0px; margin-bottom: 25px; overflow: visible; position: relative; width: 697.5px; color: #4a474b; font-family: Lato, sans-serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: 400; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="box-sizing: border-box; zoom: 1; background: #dddddd; border: none; box-shadow: none; padding: 0px; margin-bottom: 25px; overflow: visible; position: relative; width: 697.5px; color: #4a474b; font-family: Lato, sans-serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: 400; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; text-decoration-style: initial; text-decoration-color: initial;"><div class="entry clearfix" style="box-sizing: border-box; zoom: 1; padding: 0px;" data-mce-style="box-sizing: border-box; zoom: 1; padding: 0px;"><h1 class="post-title entry-title" style="box-sizing: border-box; margin: 0px; font-size: 30px; font-family: inherit; font-weight: bold; line-height: normal; color: #af1917; border: 0px none; padding: 0px; font-style: normal; overflow-wrap: break-word;" data-mce-style="box-sizing: border-box; margin: 0px; font-size: 30px; font-family: inherit; font-weight: bold; line-height: normal; color: #af1917; border: 0px none; padding: 0px; font-style: normal; overflow-wrap: break-word;">PhD Thesis position or research engineer or post-doc position in Natural Language Processing: Introduction of semantic information in a speech recognition system</h1><ul class="post-meta" style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; position: relative; font-size: 14px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; position: relative; font-size: 14px;"><li class="byline" style="box-sizing: border-box; border: none; margin: 5px 0px 0px; padding: 0px 5px 0px 0px; float: left; list-style: none; line-height: normal;" data-mce-style="box-sizing: border-box; border: none; margin: 5px 0px 0px; padding: 0px 5px 0px 0px; float: left; list-style: none; line-height: normal;"><br></li></ul><div class="entry-content clearfix" style="box-sizing: border-box; zoom: 1; clear: both; padding-top: 1.5em;" data-mce-style="box-sizing: border-box; zoom: 1; clear: both; padding-top: 1.5em;"><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;"><strong style="box-sizing: border-box; font-weight: bold;" data-mce-style="box-sizing: border-box; font-weight: bold;">Supervisors: </strong>Irina Illina, MdC, Dominique Fohr, CR CNRS</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;"><strong style="box-sizing: border-box; font-weight: bold;" data-mce-style="box-sizing: border-box; font-weight: bold;">Team:</strong> Multispeech, LORIA-INRIA</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;"><strong style="box-sizing: border-box; font-weight: bold;" data-mce-style="box-sizing: border-box; font-weight: bold;">Contact:</strong> illina@loria.fr, dominique.fohr@loria.fr</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;"><strong style="box-sizing: border-box; font-weight: bold;" data-mce-style="box-sizing: border-box; font-weight: bold;">Duration of post-doc or research engineer</strong>: 12-18 months</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;"><strong style="box-sizing: border-box; font-weight: bold;" data-mce-style="box-sizing: border-box; font-weight: bold;">Duration of PhD Thesis</strong><span> </span>: 3 years</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;"><strong style="box-sizing: border-box; font-weight: bold;" data-mce-style="box-sizing: border-box; font-weight: bold;">Deadline to apply</strong> : May 15th, 2019</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;"><strong style="box-sizing: border-box; font-weight: bold;" data-mce-style="box-sizing: border-box; font-weight: bold;">Required skills: </strong>background in statistics, natural language processing and computer program skills (Perl, Python). Candidates should email a detailed CV with diploma</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;"><br data-mce-bogus="1"></p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;">Under noisy conditions, audio acquisition is one of the toughest challenges to have a successful automatic speech recognition (ASR). Much of the success relies on the ability to attenuate ambient noise in the signal and to take it into account in the acoustic model used by the ASR. Our DNN (Deep Neural Network) denoising system and our approach to exploiting uncertainties have shown their combined effectiveness against noisy speech.</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;">The ASR stage will be supplemented by a semantic analysis. Predictive representations using continuous vectors have been shown to capture the semantic characteristics of words and their context, and to overcome representations based on counting words. Semantic analysis will be performed by combining predictive representations using continuous vectors and uncertainty on denoising. This combination will be done by the rescoring component. All our models will be based on the powerful technologies of DNN.</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;"><br></p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;"><strong style="box-sizing: border-box; font-weight: bold;" data-mce-style="box-sizing: border-box; font-weight: bold;">Main activities</strong></p><ul style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0.5em 0px 0.5em 1.5em;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0.5em 0px 0.5em 1.5em;"><li style="box-sizing: border-box; border: 0px none; margin: 0px 0px 0.5em; padding: 0px;" data-mce-style="box-sizing: border-box; border: 0px none; margin: 0px 0px 0.5em; padding: 0px;">study and implementation of a noisy speech enhancement module and a propagation of uncertainty module;</li><li style="box-sizing: border-box; border: 0px none; margin: 0px 0px 0.5em; padding: 0px;" data-mce-style="box-sizing: border-box; border: 0px none; margin: 0px 0px 0.5em; padding: 0px;">design a semantic analysis module;</li><li style="box-sizing: border-box; border: 0px none; margin: 0px 0px 0.5em; padding: 0px;" data-mce-style="box-sizing: border-box; border: 0px none; margin: 0px 0px 0.5em; padding: 0px;">design a module taking into account the semantic and uncertainty information.</li></ul><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;"><strong style="box-sizing: border-box; font-weight: bold;" data-mce-style="box-sizing: border-box; font-weight: bold;">Skills</strong></p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;">Strong background in mathematics, machine learning (DNN), statistics</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;">Following profiles are welcome, either:</p><ul style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0.5em 0px 0.5em 1.5em;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0.5em 0px 0.5em 1.5em;"><li style="box-sizing: border-box; border: 0px none; margin: 0px 0px 0.5em; padding: 0px;" data-mce-style="box-sizing: border-box; border: 0px none; margin: 0px 0px 0.5em; padding: 0px;">Strong background in signal processing</li></ul><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;">or</p><ul style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0.5em 0px 0.5em 1.5em;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0.5em 0px 0.5em 1.5em;"><li style="box-sizing: border-box; border: 0px none; margin: 0px 0px 0.5em; padding: 0px;" data-mce-style="box-sizing: border-box; border: 0px none; margin: 0px 0px 0.5em; padding: 0px;">Strong experience with natural language processing</li></ul><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;">Excellent English writing and speaking skills are required in any case.</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px;"><strong style="box-sizing: border-box; font-weight: bold;" data-mce-style="box-sizing: border-box; font-weight: bold;">References</strong></p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;">[Nathwani<span> </span><em style="box-sizing: border-box;" data-mce-style="box-sizing: border-box;">et al</em>., 2018] Nathwani, K., Vincent, E., and Illina, I. DNN uncertainty propagation using GMM-derived uncertainty features for noise robust ASR,<span> </span><em style="box-sizing: border-box;" data-mce-style="box-sizing: border-box;">IEEE Signal Processing Letters</em>, 2018.</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;">[Nathwani<span> </span><em style="box-sizing: border-box;" data-mce-style="box-sizing: border-box;">et al</em>., 2017] Nathwani, K., Vincent, E., and Illina, I. Consistent DNN uncertainty training and decoding for robust ASR, in<span> </span><em style="box-sizing: border-box;" data-mce-style="box-sizing: border-box;">Proc.</em><span> </span><em style="box-sizing: border-box;" data-mce-style="box-sizing: border-box;">IEEE Automatic Speech Recognition and Understanding Workshop</em>, 2017.</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;">[Nugraha<span> </span><em style="box-sizing: border-box;" data-mce-style="box-sizing: border-box;">et al.,</em><span> </span>2016] Nugraha, A., Liutkus, A., Vincent E. Multichannel audio source separation with deep neural networks.<span> </span><em style="box-sizing: border-box;" data-mce-style="box-sizing: border-box;">IEEE/ACM</em><span> </span><em style="box-sizing: border-box;" data-mce-style="box-sizing: border-box;">Transactions on Audio, Speech, and Language Processing</em>, 2016.</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;">[Sheikh, 2016] Sheikh, I. Exploitation du contexte sémantique pour améliorer la reconnaissance des noms propres dans les documents audio diachroniques”,<span> </span><em style="box-sizing: border-box;" data-mce-style="box-sizing: border-box;">These de doctorat en Informatique, Université de Lorraine,</em><span> </span>2016.</p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;">[Peters et al., 2017] Matthew Peters, Waleed Ammar, Chandra Bhagavatula, and Russell Power. 2017. “Semi-supervised sequence tagging with bidirectional language models.”<span> </span><em style="box-sizing: border-box;" data-mce-style="box-sizing: border-box;">In ACL.</em></p><p style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;" data-mce-style="box-sizing: border-box; margin: 0px; border: 0px none; padding: 0px; text-align: justify;">[Peters et al., 2018] Matthew Peters, Mark Neumann, Mohit Iyyer, Matt Gardner, Christopher Clark, Kenton Lee, and Luke Zettlemoyer. 2018. “Deep contextualized word representations”.<span> </span><em style="box-sizing: border-box;" data-mce-style="box-sizing: border-box;">In NAACL.</em></p></div></div></div><div class="entry-author" style="box-sizing: border-box; margin: 40px 0px; padding: 20px; border: 1px solid #ebedf0; border-radius: 5px; color: #4a474b; font-family: Lato, sans-serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: 400; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="box-sizing: border-box; margin: 40px 0px; padding: 20px; border: 1px solid #ebedf0; border-radius: 5px; color: #4a474b; font-family: Lato, sans-serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: 400; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><div class="row" style="box-sizing: border-box; margin-right: -15px; margin-left: -15px;" data-mce-style="box-sizing: border-box; margin-right: -15px; margin-left: -15px;"><div class="author-avatar col-sm-3" style="box-sizing: border-box; position: relative; min-height: 1px; padding-right: 15px; padding-left: 15px; float: left; width: 171.375px;" data-mce-style="box-sizing: border-box; position: relative; min-height: 1px; padding-right: 15px; padding-left: 15px; float: left; width: 171.375px;"><a href="https://team.inria.fr/multispeech/author/illina/" rel="author" style="box-sizing: border-box; background-color: transparent; color: #af1917; text-decoration: none;" data-mce-href="https://team.inria.fr/multispeech/author/illina/" data-mce-style="box-sizing: border-box; background-color: transparent; color: #af1917; text-decoration: none;"><img alt="" src="https://secure.gravatar.com/avatar/8a92dc68c6d376588e60aeede6ce990f?s=200&d=mm&r=g" srcset="https://secure.gravatar.com/avatar/8a92dc68c6d376588e60aeede6ce990f?s=400&d=mm&r=g 2x" class="avatar avatar-200 photo" height="200" width="200" style="box-sizing: border-box; border: 1px solid #ffffff; vertical-align: middle; margin: 0px; padding: 0px; border-radius: 100px; max-width: 100%; height: auto; box-shadow: rgba(0, 0, 0, 0.2) -1px 1px 5px;" data-mce-src="https://secure.gravatar.com/avatar/8a92dc68c6d376588e60aeede6ce990f?s=200&d=mm&r=g" data-mce-style="box-sizing: border-box; border: 1px solid #ffffff; vertical-align: middle; margin: 0px; padding: 0px; border-radius: 100px; max-width: 100%; height: auto; box-shadow: rgba(0, 0, 0, 0.2) -1px 1px 5px;"></a></div><div class="author-bio col-sm-9" style="box-sizing: border-box; position: relative; min-height: 1px; padding-right: 15px; padding-left: 15px; float: left; width: 514.125px;" data-mce-style="box-sizing: border-box; position: relative; min-height: 1px; padding-right: 15px; padding-left: 15px; float: left; width: 514.125px;"><h3 class="section-title-sm" style="box-sizing: border-box; font-family: inherit; font-weight: 500; line-height: 1.1; color: inherit; margin: 0px; font-size: 24px; border: 0px none; padding: 0px; font-style: normal;" data-mce-style="box-sizing: border-box; font-family: inherit; font-weight: 500; line-height: 1.1; color: inherit; margin: 0px; font-size: 24px; border: 0px none; padding: 0px; font-style: normal;">Irina ILLINA</h3></div></div></div><!--EndFragment--><div style="clear: both;" data-mce-style="clear: both;"><br></div></div><div><br></div><div data-marker="__SIG_POST__">-- <br></div><div>Associate Professor <br>Lorraine University<br>LORIA-INRIA<br>office C147 <br>Building C <br>615 rue du Jardin Botanique<br>54600 Villers-les-Nancy Cedex<br>Tel:+ 33 3 54 95 84 90</div></div></body></html>