<html><body><div style="font-family: times new roman, new york, times, serif; font-size: 12pt; color: #000000"><div><br></div><div style="font-family: times new roman, new york, times, serif; font-size: 12pt; color: #000000" data-mce-style="font-family: times new roman, new york, times, serif; font-size: 12pt; color: #000000;"><span style="font-size: medium;" data-mce-style="font-size: medium;"><span lang="en"><b>Post-doc position at LORIA (Nancy, France)</b></span></span><a class="mceItemAnchor" name="result_box"></a><br data-mce-bogus="1"><div><span style="font-size: medium;" data-mce-style="font-size: medium;"><span lang="en"><i><b><br></b></i></span></span></div><div><span style="font-size: medium;" data-mce-style="font-size: medium;"><span lang="en"><i><b>Automatic speech recognition: Deep Neural Network for <span style="font-size: medium;" data-mce-style="font-size: medium;">L</span>anguage <span style="font-size: medium;" data-mce-style="font-size: medium;">M</span>odel </b></i></span></span></div><div><div style="font-family: times new roman, new york, times, serif; font-size: 12pt; color: #000000" data-mce-style="font-family: times new roman, new york, times, serif; font-size: 12pt; color: #000000;"><div><div style="font-family: times new roman, new york, times, serif; font-size: 12pt; color: #000000" data-mce-style="font-family: times new roman, new york, times, serif; font-size: 12pt; color: #000000;"><div><p style="margin-bottom: 0cm" data-mce-style="margin-bottom: 0cm;" align="JUSTIFY"><span lang="en"><b>Framework of ANR project </b></span><span lang="en"><i><b>ContNomina </b></i></span></p><p style="margin-bottom: 0cm" data-mce-style="margin-bottom: 0cm;" align="JUSTIFY"><span lang="en">The technologies involved in information retrieval in large audio/video databases are often based on the analysis of large, but closed, corpora, and on machine learning techniques and statistical modeling of the written and spoken language. The effectiveness of these approaches is now widely acknowledged, but they nevertheless have major flaws, particularly for what concern proper names, that are crucial for the interpretation of the content.</span></p><p style="margin-bottom: 0cm" data-mce-style="margin-bottom: 0cm;" align="JUSTIFY"><span lang="en">In the context of diachronic data (data which change over time) new proper names appear constantly requiring dynamic updates of the lexicons and language models used by the speech recognition system.</span></p><p style="margin-bottom: 0cm" data-mce-style="margin-bottom: 0cm;" align="JUSTIFY"><span lang="en">As a result, the ANR project </span><span lang="en"><i>ContNomina</i></span><span lang="en"> (2013-2017) focuses on the problem of proper names in automatic audio processing systems by exploiting in the most efficient way the context of the processed documents. To do this, the post-doc student will address </span><span style="color: #333333;" data-mce-style="color: #333333;"><span lang="en">the contextualization of the recognition module through the dynamic adjustment of the language model in order to make it more accurate.</span></span></p><p style="margin-bottom: 0cm" data-mce-style="margin-bottom: 0cm;" align="JUSTIFY"><span style="color: #333333;" data-mce-style="color: #333333;"><span lang="en"><b>Post-doc subject</b></span></span></p><p style="margin-bottom: 0cm" data-mce-style="margin-bottom: 0cm;" align="JUSTIFY"><span style="color: #333333;" data-mce-style="color: #333333;"><span lang="en"><b>Deep Neural Network</b></span></span><span style="color: #333333;" data-mce-style="color: #333333;"><span lang="en"> have become a key component of modern automatic speech recognition systems. The language model of our recognition system is based on a neural network learned from a large corpus of text. The problem is to estimate the probability of a new proper name depending on its context. Several tracks will be explored: adapting the language model, using a class model or studying the notion of analogy.</span></span></p><p style="margin-bottom: 0cm" data-mce-style="margin-bottom: 0cm;" align="JUSTIFY"><span style="color: #333333;" data-mce-style="color: #333333;"><span lang="en">Our team has developed a fully automatic system for speech recognition to transcribe a radio broadcast from the corresponding audio file. The post-doc will develop a new module whose function is to integrate new proper names in the language model.</span></span></p><p style="margin-bottom: 0cm" data-mce-style="margin-bottom: 0cm;" align="JUSTIFY"><span style="color: #333333;" data-mce-style="color: #333333;"><span lang="en"><b>Required skills</b></span></span></p><p style="margin-bottom: 0cm" data-mce-style="margin-bottom: 0cm;" align="JUSTIFY"><span style="color: #333333;" data-mce-style="color: #333333;"><span lang="en">A PhD </span></span><span style="color: #333333;" data-mce-style="color: #333333;"><span style="font-family: Calibri,sans-serif;" data-mce-style="font-family: Calibri,sans-serif;"><span lang="en">in signal processing or in computer sciences</span></span></span><span style="color: #333333;" data-mce-style="color: #333333;"><span lang="en">, be familiar with the tools for automatic speech recognition, background in statistics and computer program skills (C, object-oriented programming and Perl).</span></span></p><p style="margin-bottom: 0cm" data-mce-style="margin-bottom: 0cm;" align="JUSTIFY"><a class="mceItemAnchor" name="__DdeLink__72_1088968567"></a> <span style="color: #333333;" data-mce-style="color: #333333;"><span lang="en-US"><b>Post-doc duration: </b></span></span><span style="color: #333333;" data-mce-style="color: #333333;"><span lang="en-US">12 months, start during</span></span><span lang="en-US"> the 2015 (these is some flexibility) </span></p><p style="margin-bottom: 0cm" data-mce-style="margin-bottom: 0cm;"><span lang="en-US"><b>Localization and contacts: </b></span><span lang="en-US">Loria laboratory, </span><span lang="en-US"><i>Speech team</i></span><span lang="en-US">, Nancy, France</span></p><p style="margin-bottom: 0cm" data-mce-style="margin-bottom: 0cm;"><span style="color: #0000ff;" data-mce-style="color: #0000ff;"><span lang="en-US"><span style="text-decoration: underline;" data-mce-style="text-decoration: underline;"><span lang="en-US">i<span class="Object" id="OBJ_PREFIX_DWT4518_com_zimbra_email"><span class="Object" id="OBJ_PREFIX_DWT4524_com_zimbra_email"><a href="mailto:Irina.illina@loria.fr" target="_blank" data-mce-href="mailto:Irina.illina@loria.fr">rina.illina@loria.fr</a></span></span></span></span></span></span><span lang="en-US"> </span><span style="color: #0000ff;" data-mce-style="color: #0000ff;"><span lang="en-US"><span style="text-decoration: underline;" data-mce-style="text-decoration: underline;"><span lang="en-US"><span class="Object" id="OBJ_PREFIX_DWT4519_com_zimbra_email"><span class="Object" id="OBJ_PREFIX_DWT4525_com_zimbra_email"><a href="mailto:dominique.fohr@loria.fr" target="_blank" data-mce-href="mailto:dominique.fohr@loria.fr">dominique.fohr@loria.fr</a></span></span> <span class="Object" id="OBJ_PREFIX_DWT4520_com_zimbra_email"><span class="Object" id="OBJ_PREFIX_DWT4526_com_zimbra_email"><a href="mailto:georges.linares@univ-avignon.fr" target="_blank" data-mce-href="mailto:georges.linares@univ-avignon.fr">georges.linares@univ-avignon.fr</a></span></span></span></span></span></span><br></p><p style="margin-bottom: 0cm" data-mce-style="margin-bottom: 0cm;"><span lang="en-US">Candidates should email a letter of application, a detailed CV with a list of publications and diploma</span></p><br></div><div><br></div><div>Irina Illina <br></div><div>-- <br></div><div><span></span>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<span></span><br></div></div></div><div><br></div></div></div></div></div></body></html>