{"id":26150,"date":"2024-01-18T15:05:02","date_gmt":"2024-01-18T14:05:02","guid":{"rendered":"https:\/\/lium.univ-lemans.fr\/?p=26150"},"modified":"2024-07-01T14:37:24","modified_gmt":"2024-07-01T12:37:24","slug":"soutenance-de-these-valentin-pelloin","status":"publish","type":"post","link":"https:\/\/lium.univ-lemans.fr\/en\/soutenance-de-these-valentin-pelloin\/","title":{"rendered":"Valentin Pelloin"},"content":{"rendered":"<div class=\"panel-grid\" id=\"pg-26150-0\" ><div class=\"panel-grid-core\"><div class=\"panel-grid-cell\" id=\"pgc-26150-0-0\" ><div class=\"panel-widget-style\" ><h2 style=\"text-align: center;\"><span style=\"color: #e5442d;\">PhD defence,  Valentin Pelloin <\/span><\/h2>\n<p><b>Date : <\/b>24\/01\/2024<br \/>\n<b>Time : <\/b> 9h30<br \/>\n<b>Location : <\/b> Le Mans Universit\u00e9; IC2 buiding Auditorium<br \/>\n&nbsp;<\/p>\n<p><span style=\"font-size: 18pt;\"><strong style=\"color: #e5442d;\">Title: Spoken language understanding in human-computer dialogue systems in the era of pre-trained models<\/strong><\/span><br \/>\n&nbsp;<\/p>\n<p><span style=\"color: #e5442d;\"><strong><span style=\"font-size: 14pt;\">Jury members :<\/span><br \/>\n<\/strong><\/span><span style=\"font-size: 12pt;\"><\/p>\n<ul>\n<li><strong>Christophe CERISARA<\/strong>, Researcher, LORIA, Nancy, <strong>Reviewer <\/strong><\/li>\n<li><strong>Benoit FAVRE<\/strong>, Professor, LIS, Marseille, <strong>Reviewer <\/strong><\/li>\n<li><strong>G\u00e9raldine DAMNATI<\/strong>, Research Engineer, Orange Labs, Lannion <strong>Examiner <\/strong><\/li>\n<li><strong>Richard DUFOUR<\/strong>, Assistant Professor, LIA, Avignon,  <strong>Examiner <\/strong><\/li>\n<li><strong>Sophie ROSSET<\/strong>, Research Director, LISN, Orsay <strong>Examiner <\/strong><\/li>\n<li><strong>Sylvain MEIGNIER<\/strong>, Professor, Le Mans Universit\u00e9 LIUM, <strong>Director of thesis <\/strong><\/li>\n<li><strong>Nathalie CAMELIN<\/strong>, Assistant Professor, Le Mans Universit\u00e9 LIUM, <strong>Supervisor <\/strong><\/li>\n<li><strong>Antoine LAURENT<\/strong>, Professor, Le Mans Universit\u00e9 LIUM, <strong>Supervisor <\/strong><\/li>\n<\/ul>\n<p><\/span><\/p>\n<p>&nbsp;<\/p>\n<p><span style=\"color: #e5442d;\"><strong><span style=\"font-size: 14pt;\">Abstract:<br \/>\n<\/span> <\/strong><\/span><\/p>\n<p style=\"text-align: justify;\" align=\"justify\"> In this thesis, spoken language understanding (SLU) is studied in the application context of telephone dialogues with defined goals (hotel booking reservations, for example).<\/p>\n<p style=\"text-align: justify;\" align=\"justify\">Historically, SLU was performed through a cascade of systems: a first system would transcribe the speech into words, and a natural language understanding system would link those words to a semantic annotation. The development of deep neural methods has led to the emergence of end-to-end architectures, where the understanding task is performed by a single system, applied directly to the speech signal to extract the semantic annotation. <\/p>\n<p style=\"text-align: justify;\" align=\"justify\">Recently, so-called self-supervised learning (SSL) pre-trained models have brought new advances in natural language processing (NLP). Learned in a generic way on very large datasets, they can then be adapted for other applications. To date, the best SLU results have been obtained with pipeline systems incorporating SSL models.<\/p>\n<p style=\"text-align: justify;\" align=\"justify\">However, none of the architectures, pipeline or end-to-end, is perfect. In this thesis, we study these architectures and propose hybrid versions that attempt to benefit from the advantages of each. After developing a state-of-the-art end-to-end SLU model, we evaluated different hybrid strategies. The advances made by SSL models during the course of this thesis led us to integrate them into our hybrid architecture.<\/p>\n<p>&nbsp;<\/p>\n<p><span style=\"color: #e5442d;\"><strong><span style=\"font-size: 14pt;\">Keywords:<br \/>\n<\/span> <\/strong><\/span><\/p>\n<p style=\"text-align: justify;\" align=\"justify\">  spoken language understanding, automatic speech recognition, neural networks, pre-trained models, self-supervised models, semantic concepts extraction<\/p><\/div><\/div><\/div><\/div><div class=\"panel-grid\" id=\"pg-26150-1\" ><div class=\"panel-grid-core\"><div class=\"panel-grid-cell\" id=\"pgc-26150-1-0\" ><div class=\"panel-widget-style\" ><div class=\"margin10\"><\/div><\/div><\/div><\/div><\/div><div class=\"panel-grid\" id=\"pg-26150-2\" ><div class=\"panel-grid-core\"><div class=\"panel-grid-cell\" id=\"pgc-26150-2-0\" >&nbsp;<\/div><div class=\"panel-grid-cell\" id=\"pgc-26150-2-1\" >&nbsp;<\/div><div class=\"panel-grid-cell\" id=\"pgc-26150-2-2\" >&nbsp;<\/div><\/div><\/div><div class=\"panel-grid\" id=\"pg-26150-3\" ><div class=\"panel-grid-core\"><div class=\"panel-grid-cell\" id=\"pgc-26150-3-0\" ><div class=\"panel-widget-style\" ><div class=\"margin10\"><\/div><\/div><\/div><\/div><\/div><div class=\"panel-grid\" id=\"pg-26150-4\" ><div class=\"panel-grid-core\"><div class=\"panel-grid-cell\" id=\"pgc-26150-4-0\" >&nbsp;<\/div><div class=\"panel-grid-cell\" id=\"pgc-26150-4-1\" ><div class=\"panel-widget-style\" ><p><img src=\"https:\/\/lium.univ-lemans.fr\/wp-content\/uploads\/2020\/12\/logo_LEMANS_UNIVERSITE-01ptFormat.jpg\" alt=\"\" \/ ><\/p><\/div><\/div><div class=\"panel-grid-cell\" id=\"pgc-26150-4-2\" >&nbsp;<\/div><div class=\"panel-grid-cell\" id=\"pgc-26150-4-3\" ><div class=\"panel-widget-style\" ><p><img src=\"https:\/\/lium.univ-lemans.fr\/wp-content\/uploads\/2023\/09\/Mathematiques-STIC-ED.png\" alt=\"\" \/ ><\/p><\/div><\/div><div class=\"panel-grid-cell\" id=\"pgc-26150-4-4\" >&nbsp;<\/div><\/div><\/div>","protected":false},"excerpt":{"rendered":"<p>PhD defence, Valentin Pelloin Date : 24\/01\/2024 Time : 9h30 Location : Le Mans Universit\u00e9; IC2 buiding Auditorium &nbsp; Title: Spoken language understanding in human-computer dialogue systems in the era of pre-trained models &nbsp; Jury members : Christophe CERISARA, Researcher, LORIA, Nancy, Reviewer Benoit FAVRE, Professor, LIS, Marseille, Reviewer G\u00e9raldine DAMNATI, Research Engineer, Orange Labs, [&hellip;]<\/p>\n<p class=\"more-link style2\"><a href=\"https:\/\/lium.univ-lemans.fr\/en\/soutenance-de-these-valentin-pelloin\/\"  class=\"themebutton\"><span class=\"more-text\">READ MORE<\/span><span class=\"more-icon\"><i class=\"fa fa-angle-right fa-lg\"><\/i><\/span><\/a><\/p>\n","protected":false},"author":14,"featured_media":24068,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[46,6],"tags":[49],"acf":[],"_links":{"self":[{"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/posts\/26150"}],"collection":[{"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/users\/14"}],"replies":[{"embeddable":true,"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/comments?post=26150"}],"version-history":[{"count":0,"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/posts\/26150\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/media\/24068"}],"wp:attachment":[{"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/media?parent=26150"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/categories?post=26150"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/tags?post=26150"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}