{"id":337,"date":"2012-10-01T10:43:04","date_gmt":"2012-10-01T08:43:04","guid":{"rendered":"https:\/\/samovar2022.int-evry.fr\/index.php\/2012\/10\/01\/soutenance-theses-de-selwa-rafi\/"},"modified":"2020-09-04T18:46:58","modified_gmt":"2020-09-04T16:46:58","slug":"soutenance-theses-de-selwa-rafi","status":"publish","type":"post","link":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/2012\/10\/01\/soutenance-theses-de-selwa-rafi\/","title":{"rendered":"SOUTENANCE : Th\u00e8ses de Selwa Rafi"},"content":{"rendered":"

\u00abCha\u00eenes de Markov cach\u00e9es et s\u00e9paration non supervis\u00e9e de sources\u00bb<\/strong>.<\/p>\n

lundi 11 juin<\/strong> 2012 \u00e0 14h30 en salle C06 – 9 rue Charles Fourier, 91000 Evry. <\/p>\n

Jury :<\/strong><\/p>\n

— M. Ali Mohammad-Djafari, DR-CNRS \u00e0 Sup\u00e9lec-Univ. Paris-Sud (rapporteur)<\/p>\n

— M. Christian Jutten, Professeur \u00e0 l’Universit\u00e9 Grenoble I (rapporteur)<\/p>\n

— M. Pierre Comon, DR-CNRS \u00e0 I3S Sophia Antipolis <\/p>\n

— M. Yannick Deville, Professeur \u00e0 l’Universit\u00e9 Toulouse 3<\/p>\n

— M. Michel Broniatowski, Professeur \u00e0 l’Universit\u00e9 Paris 6<\/p>\n

— M. Wojciech Pieczynski, Professeur \u00e0 T\u00e9l\u00e9com SudParis (directeur de th\u00e8se) <\/p>\n

— M. Marc Castella, Ma\u00eetre de Conf\u00e9rences \u00e0 T\u00e9l\u00e9com SudParis (co-encadrant)<\/p>\n

R\u00e9sum\u00e9<\/strong><\/p>\n

Le probl\u00e8me de la restauration est rencontr\u00e9 dans domaines tr\u00e8s vari\u00e9s notamment en traitement de signal et de l’image. Il correspond \u00e0 la r\u00e9cup\u00e9ration des donn\u00e9es originales \u00e0 partir de donn\u00e9es observ\u00e9es. Dans le cas de donn\u00e9es multidimensionnelles, la r\u00e9solution de ce probl\u00e8me peut se faire par diff\u00e9rentes approches selon la nature des donn\u00e9es, l\u2019op\u00e9rateur de transformation et la pr\u00e9sence ou non de bruit.<\/p>\n

Dans ce travail, nous avons trait\u00e9 ce probl\u00e8me, d\u2019une part, dans le cas de donn\u00e9es discr\u00e8tes en pr\u00e9sence de bruit. Dans ce cas, le probl\u00e8me de restauration est analogue \u00e0 celui de la segmentation. Nous avons alors exploit\u00e9 les mod\u00e9lisations dites cha\u00eenes de Markov couples et triplets qui g\u00e9n\u00e9ralisent les cha\u00eenes de Markov cach\u00e9es. L’int\u00e9r\u00eat de ces mod\u00e8les r\u00e9side en la possibilit\u00e9 de g\u00e9n\u00e9raliser la m\u00e9thode de calcul de la probabilit\u00e9 a posteriori, ce qui permet une segmentation bay\u00e9sienne. Nous avons consid\u00e9r\u00e9 ces m\u00e9thodes pour des observations bi-dimensionnelles et nous avons appliqu\u00e9 les algorithmes pour une s\u00e9paration sur des documents issus de manuscrits scann\u00e9s dans lesquels les textes des deux faces d\u2019une feuille se m\u00e9langeaient.<\/p>\n

D\u2019autre part, nous avons attaqu\u00e9 le probl\u00e8me de la restauration dans un contexte de s\u00e9paration aveugle de sources. Une m\u00e9thode classique en s\u00e9paration aveugle de sources, connue sous l\u2019appellation \u00ab\u00a0Analyse en Composantes Ind\u00e9pendantes\u00a0\u00bb (ACI), n\u00e9cessite l’hypoth\u00e8se d’ind\u00e9pendance statistique des sources. Dans des situations r\u00e9elles, cette hypoth\u00e8se n\u2019est pas toujours v\u00e9rifi\u00e9e. Par cons\u00e9quent, nous avons \u00e9tudi\u00e9 une extension du mod\u00e8le ACI dans le cas o\u00f9 les sources peuvent \u00eatre statistiquement d\u00e9pendantes. Pour ce faire, nous avons introduit un processus latent qui gouverne la d\u00e9pendance et\/ou l\u2019ind\u00e9pendance des sources. Le mod\u00e8le que nous proposons combine un mod\u00e8le de m\u00e9lange lin\u00e9aire instantan\u00e9 tel que celui donn\u00e9 par ACI et un mod\u00e8le probabiliste sur les sources avec variables cach\u00e9es. Dans ce cadre, nous montrons comment la technique d’Estimation Conditionnelle It\u00e9rative permet d’affaiblir l’hypoth\u00e8se usuelle d’ind\u00e9pendance en une hypoth\u00e8se d’ind\u00e9pendance conditionnelle.<\/p>\n

Mots-Cl\u00e9s:<\/strong> Mod\u00e8les de Markov cach\u00e9s, Mod\u00e8les de Markov Couples et Triplets, Estimation Bay\u00e9sienne, S\u00e9paration aveugle de sources, Analyse en Composantes Ind\u00e9pendantes, Estimation Conditionnelle It\u00e9rative<\/p>\n

\n","protected":false},"excerpt":{"rendered":"

\u00abCha\u00eenes de Markov cach\u00e9es et s\u00e9paration non supervis\u00e9e de sources\u00bb. lundi 11 juin 2012 \u00e0 14h30 en salle C06 – 9 rue Charles Fourier, 91000 Evry. Jury : — M. Ali Mohammad-Djafari, DR-CNRS \u00e0 Sup\u00e9lec-Univ. Paris-Sud (rapporteur) — M. Christian Jutten, Professeur \u00e0 l’Universit\u00e9 Grenoble I (rapporteur) — M. Pierre Comon, DR-CNRS \u00e0 I3S Sophia […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"ocean_post_layout":"","ocean_both_sidebars_style":"","ocean_both_sidebars_content_width":0,"ocean_both_sidebars_sidebars_width":0,"ocean_sidebar":"","ocean_second_sidebar":"","ocean_disable_margins":"enable","ocean_add_body_class":"","ocean_shortcode_before_top_bar":"","ocean_shortcode_after_top_bar":"","ocean_shortcode_before_header":"","ocean_shortcode_after_header":"","ocean_has_shortcode":"","ocean_shortcode_after_title":"","ocean_shortcode_before_footer_widgets":"","ocean_shortcode_after_footer_widgets":"","ocean_shortcode_before_footer_bottom":"","ocean_shortcode_after_footer_bottom":"","ocean_display_top_bar":"default","ocean_display_header":"default","ocean_header_style":"","ocean_center_header_left_menu":"","ocean_custom_header_template":"","ocean_custom_logo":0,"ocean_custom_retina_logo":0,"ocean_custom_logo_max_width":0,"ocean_custom_logo_tablet_max_width":0,"ocean_custom_logo_mobile_max_width":0,"ocean_custom_logo_max_height":0,"ocean_custom_logo_tablet_max_height":0,"ocean_custom_logo_mobile_max_height":0,"ocean_header_custom_menu":"","ocean_menu_typo_font_family":"","ocean_menu_typo_font_subset":"","ocean_menu_typo_font_size":0,"ocean_menu_typo_font_size_tablet":0,"ocean_menu_typo_font_size_mobile":0,"ocean_menu_typo_font_size_unit":"px","ocean_menu_typo_font_weight":"","ocean_menu_typo_font_weight_tablet":"","ocean_menu_typo_font_weight_mobile":"","ocean_menu_typo_transform":"","ocean_menu_typo_transform_tablet":"","ocean_menu_typo_transform_mobile":"","ocean_menu_typo_line_height":0,"ocean_menu_typo_line_height_tablet":0,"ocean_menu_typo_line_height_mobile":0,"ocean_menu_typo_line_height_unit":"","ocean_menu_typo_spacing":0,"ocean_menu_typo_spacing_tablet":0,"ocean_menu_typo_spacing_mobile":0,"ocean_menu_typo_spacing_unit":"","ocean_menu_link_color":"","ocean_menu_link_color_hover":"","ocean_menu_link_color_active":"","ocean_menu_link_background":"","ocean_menu_link_hover_background":"","ocean_menu_link_active_background":"","ocean_menu_social_links_bg":"","ocean_menu_social_hover_links_bg":"","ocean_menu_social_links_color":"","ocean_menu_social_hover_links_color":"","ocean_disable_title":"default","ocean_disable_heading":"default","ocean_post_title":"","ocean_post_subheading":"","ocean_post_title_style":"","ocean_post_title_background_color":"","ocean_post_title_background":0,"ocean_post_title_bg_image_position":"","ocean_post_title_bg_image_attachment":"","ocean_post_title_bg_image_repeat":"","ocean_post_title_bg_image_size":"","ocean_post_title_height":0,"ocean_post_title_bg_overlay":0.5,"ocean_post_title_bg_overlay_color":"","ocean_disable_breadcrumbs":"default","ocean_breadcrumbs_color":"","ocean_breadcrumbs_separator_color":"","ocean_breadcrumbs_links_color":"","ocean_breadcrumbs_links_hover_color":"","ocean_display_footer_widgets":"default","ocean_display_footer_bottom":"default","ocean_custom_footer_template":"","ocean_post_oembed":"","ocean_post_self_hosted_media":"","ocean_post_video_embed":"","ocean_link_format":"","ocean_link_format_target":"self","ocean_quote_format":"","ocean_quote_format_link":"post","ocean_gallery_link_images":"on","ocean_gallery_id":[],"footnotes":""},"categories":[418],"tags":[],"class_list":["post-337","post","type-post","status-publish","format-standard","hentry","category-theses-2012-fr","entry"],"_links":{"self":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/337","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/comments?post=337"}],"version-history":[{"count":1,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/337\/revisions"}],"predecessor-version":[{"id":1887,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/337\/revisions\/1887"}],"wp:attachment":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/media?parent=337"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/categories?post=337"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/tags?post=337"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}