{"id":12847,"date":"2025-08-07T12:00:52","date_gmt":"2025-08-07T12:00:52","guid":{"rendered":"https:\/\/ingch.ch\/?p=12847"},"modified":"2025-08-07T13:59:30","modified_gmt":"2025-08-07T13:59:30","slug":"comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains","status":"publish","type":"post","link":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/","title":{"rendered":"Comment faire pour que l’IA voie aussi bien que les humains"},"content":{"rendered":"

Comment faire pour que l’IA voie aussi bien que les humains<\/strong>
\nUne \u00e9tude de l\u2019EPFL r\u00e9v\u00e8le pourquoi les \u00eatres humains excellent dans la reconnaissance des objets \u00e0 partir de fragments alors que l\u2019IA peine \u00e0 le faire, soulignant le r\u00f4le crucial de l\u2019int\u00e9gration des contours dans la vision humaine.<\/em><\/p>\n

Auteur: Nik Papageorgiou
\n<\/em>Source: EPFL<\/a>
\nImage: (EPFL\/iStock (Vertigo3d)<\/em><\/p>\n

Chaque jour, nous reconnaissons sans effort des amis dans une foule ou identifions des formes famili\u00e8res, m\u00eame si elles sont partiellement cach\u00e9es. Notre cerveau assemble des fragments pour en faire des objets entiers, remplissant les blancs pour donner un sens \u00e0 un monde souvent chaotique.<\/p>\n

Cette capacit\u00e9 est appel\u00e9e \u00abint\u00e9gration des contours\u00bb. M\u00eame les syst\u00e8mes d\u2019IA les plus intelligents ont encore du mal \u00e0 effectuer cette t\u00e2che. Malgr\u00e9 les r\u00e9alisations remarquables de l\u2019intelligence artificielle dans la reconnaissance d\u2019images, les IA peinent encore \u00e0 g\u00e9n\u00e9raliser \u00e0 partir d\u2019informations visuelles incompl\u00e8tes ou fragment\u00e9es.<\/p>\n

Lorsque les objets sont partiellement cach\u00e9s, effac\u00e9s ou fragment\u00e9s, la plupart des mod\u00e8les d\u2019IA h\u00e9sitent, classent incorrectement ou abandonnent. Cela peut \u00eatre un grave probl\u00e8me dans la vie r\u00e9elle, compte tenu de notre d\u00e9pendance croissante \u00e0 l\u2019IA pour des applications, telles que les voitures autonomes, les proth\u00e8ses et la robotique.Anchor<\/p>\n

L’\u00e9quipe du NeuroAI Lab de l\u2019EPFL, dirig\u00e9e par Martin Schrimpf, a entrepris de comparer syst\u00e9matiquement la mani\u00e8re dont les personnes et l\u2019IA g\u00e8rent les puzzles visuels. Ben Lonnqvist, \u00e9tudiant dipl\u00f4m\u00e9 de l\u2019EDNE et principal auteur de l\u2019\u00e9tude, a collabor\u00e9 avec le Laboratoire de psychophysique du professeur Michael Herzog pour d\u00e9velopper une s\u00e9rie de tests de reconnaissance dans lesquels des \u00eatres humains et plus de 1000 r\u00e9seaux neuronaux artificiels devaient identifier des objets dont les contours \u00e9taient manquants ou fragment\u00e9s. Leurs r\u00e9sultats montrent qu\u2019en mati\u00e8re d\u2019int\u00e9gration des contours, les \u00eatres humains surpassent syst\u00e9matiquement l\u2019IA de pointe, et r\u00e9v\u00e8lent pourquoi.<\/p>\n

La recherche a \u00e9t\u00e9 pr\u00e9sent\u00e9e \u00e0 la Conf\u00e9rence internationale sur l\u2019apprentissage machine (ICML) de 2025<\/a>.<\/p>\n

\n
\"Beispiel\"
Les humains peuvent identifier des objets \u00e0 partir de donn\u00e9es partielles, tandis que les IA ont du mal \u00e0 le faire. Cr\u00e9dit : Ben L\u00f6nnvist (EPFL) <\/em><\/figcaption><\/figure>\n<\/div>\n
\n

Des tests de plus en plus difficiles<\/strong><\/p>\n

L\u2019\u00e9quipe a mis en place un test de reconnaissance d\u2019objets en laboratoire avec une cinquantaine de volontaires. Les participantes et participants ont visualis\u00e9 des images d\u2019objets du quotidien, tels que des tasses, des chapeaux et des casseroles, dont les contours \u00e9taient syst\u00e9matiquement effac\u00e9s ou d\u00e9compos\u00e9s en segments. Parfois, seulement 35 % des contours d\u2019un objet restaient visibles. En parall\u00e8le, l\u2019\u00e9quipe a confi\u00e9 la m\u00eame t\u00e2che \u00e0 plus de 1000 mod\u00e8les d\u2019IA, dont certains des syst\u00e8mes les plus puissants disponibles.<\/p>\n

L\u2019exp\u00e9rience a port\u00e9 sur 20 conditions diff\u00e9rentes, variant le type et la quantit\u00e9 d\u2019informations visuelles. L\u2019\u00e9quipe a compar\u00e9 les performances dans ces conditions, mesurant la pr\u00e9cision et analysant la mani\u00e8re dont les \u00eatres humains et les machines r\u00e9agissaient \u00e0 des puzzles visuels de plus en plus difficiles.<\/p>\n

Les \u00eatres humains se sont av\u00e9r\u00e9s remarquablement performants, obtenant souvent une pr\u00e9cision de 50 % m\u00eame lorsque la plus grande partie du contour d\u2019un objet \u00e9tait absente. Les mod\u00e8les d\u2019IA, en revanche, avaient tendance \u00e0 se perdre dans des suppositions al\u00e9atoires dans les m\u00eames conditions. Seuls les mod\u00e8les entra\u00een\u00e9s sur des milliards d\u2019images se sont rapproch\u00e9s des performances humaines et, m\u00eame alors, ils ont d\u00fb \u00eatre sp\u00e9cifiquement adapt\u00e9s aux images de l\u2019\u00e9tude.<\/p>\n

En creusant davantage, les chercheuses et chercheurs ont d\u00e9couvert que les \u00eatres humains pr\u00e9f\u00e8rent naturellement reconna\u00eetre des objets lorsque des parties fragment\u00e9es pointent dans la m\u00eame direction, ce que l\u2019\u00e9quipe a appel\u00e9 le \u00abbiais d\u2019int\u00e9gration\u00bb. Les mod\u00e8les d\u2019IA entra\u00een\u00e9s pour d\u00e9velopper un biais similaire ont obtenu de meilleurs r\u00e9sultats lorsqu\u2019ils \u00e9taient confront\u00e9s \u00e0 des distorsions d\u2019image. L\u2019entra\u00eenement des syst\u00e8mes d\u2019IA sp\u00e9cialement con\u00e7us pour int\u00e9grer les contours a am\u00e9lior\u00e9 leur pr\u00e9cision et leur a \u00e9galement permis de se concentrer davantage sur la forme d\u2019un objet que sur sa texture.<\/p>\n

Donner aux machines un contenu plus humain<\/p>\n

Ces r\u00e9sultats sugg\u00e8rent que l\u2019int\u00e9gration des contours n\u2019est pas une caract\u00e9ristique inn\u00e9e, mais qu\u2019elle peut \u00eatre apprise de l\u2019exp\u00e9rience. Pour les industries qui s\u2019appuient sur la vision par ordinateur, comme les voitures autonomes ou l\u2019imagerie m\u00e9dicale, cr\u00e9er une IA qui voit le monde davantage comme nous pourrait aboutir \u00e0 une technologie plus s\u00fbre et plus fiable.<\/p>\n

Les travaux montrent \u00e9galement que le meilleur moyen de combler cet \u00e9cart n\u2019est pas de bricoler avec des architectures d\u2019IA, mais de donner aux machines un contenu visuel plus \u00abhumain\u00bb, comprenant plusieurs images du monde r\u00e9el o\u00f9 les objets sont souvent partiellement cach\u00e9s.<\/p>\n

 <\/p>\n<\/div>\n

R\u00e9f\u00e9rences<\/strong>\r\nBen L\u00f6nnqvist, Elsa Scialom, Abdulkadir Gokce, Zehra Merchant, Michael H. Herzog, Martin Schrimpf. Contour Integration Underlies Human-Like Vision. arXiv: 2504.05253<\/a><\/pre>\n","protected":false},"excerpt":{"rendered":"
Comment faire pour que l’IA voie aussi bien que les humains Une \u00e9tude de l\u2019EPFL r\u00e9v\u00e8le pourquoi les \u00eatres humains… Continue Reading Comment faire pour que l’IA voie aussi bien que les humains<\/span><\/a><\/p>\n","protected":false},"author":2,"featured_media":12853,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"inline_featured_image":false,"footnotes":""},"categories":[13,14,55],"tags":[94,214,269,270],"class_list":["post-12847","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ingflash","category-news","category-science-technology","tag-nachwuchsfoerderung","tag-technik","tag-frauen","tag-foerderung"],"yoast_head":"\nIngCH<\/title>\n<meta name=\"description\" content=\"Une \u00e9tude de l\u2019EPFL r\u00e9v\u00e8le pourquoi les \u00eatres humains excellent dans la reconnaissance des objets \u00e0 partir de fragments alors que l\u2019IA peine\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"IngCH\" \/>\n<meta property=\"og:description\" content=\"Une \u00e9tude de l\u2019EPFL r\u00e9v\u00e8le pourquoi les \u00eatres humains excellent dans la reconnaissance des objets \u00e0 partir de fragments alors que l\u2019IA peine\" \/>\n<meta property=\"og:url\" content=\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/\" \/>\n<meta property=\"og:site_name\" content=\"IngCH\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/facebook.com\/ingch2\/\" \/>\n<meta property=\"article:published_time\" content=\"2025-08-07T12:00:52+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2025-08-07T13:59:30+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/688x387-1.jpg\" \/>\n\t<meta property=\"og:image:width\" content=\"688\" \/>\n\t<meta property=\"og:image:height\" content=\"386\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/jpeg\" \/>\n<meta name=\"author\" content=\"lh\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"lh\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"5 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/\"},\"author\":{\"name\":\"lh\",\"@id\":\"https:\/\/ingch.ch\/en\/#\/schema\/person\/81354b74b93673fafb73b27fb31de962\"},\"headline\":\"Comment faire pour que l’IA voie aussi bien que les humains\",\"datePublished\":\"2025-08-07T12:00:52+00:00\",\"dateModified\":\"2025-08-07T13:59:30+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/\"},\"wordCount\":887,\"publisher\":{\"@id\":\"https:\/\/ingch.ch\/en\/#organization\"},\"image\":{\"@id\":\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/688x387-1.jpg\",\"keywords\":[\"Nachwuchsf\u00f6rderung\",\"Technik\",\"Frauen\",\"F\u00f6rderung\"],\"articleSection\":[\"IngFLASH\",\"News\",\"Science & Technology\"],\"inLanguage\":\"en-US\"},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/\",\"url\":\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/\",\"name\":\"IngCH\",\"isPartOf\":{\"@id\":\"https:\/\/ingch.ch\/en\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#primaryimage\"},\"image\":{\"@id\":\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/688x387-1.jpg\",\"datePublished\":\"2025-08-07T12:00:52+00:00\",\"dateModified\":\"2025-08-07T13:59:30+00:00\",\"description\":\"Une \u00e9tude de l\u2019EPFL r\u00e9v\u00e8le pourquoi les \u00eatres humains excellent dans la reconnaissance des objets \u00e0 partir de fragments alors que l\u2019IA peine\",\"breadcrumb\":{\"@id\":\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#primaryimage\",\"url\":\"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/688x387-1.jpg\",\"contentUrl\":\"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/688x387-1.jpg\",\"width\":688,\"height\":386,\"caption\":\"(c) EPFL\/iStock (Vertigo3d)\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Startseite\",\"item\":\"https:\/\/ingch.ch\/en\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Comment faire pour que l’IA voie aussi bien que les humains\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/ingch.ch\/en\/#website\",\"url\":\"https:\/\/ingch.ch\/en\/\",\"name\":\"IngCH\",\"description\":\"MINT for our future\",\"publisher\":{\"@id\":\"https:\/\/ingch.ch\/en\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/ingch.ch\/en\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/ingch.ch\/en\/#organization\",\"name\":\"IngCH Engineers Shape our Future\",\"url\":\"https:\/\/ingch.ch\/en\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/ingch.ch\/en\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/2021\/06\/ingch_logo-RGB.svg\",\"contentUrl\":\"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/2021\/06\/ingch_logo-RGB.svg\",\"width\":\"156\",\"height\":\"87\",\"caption\":\"IngCH Engineers Shape our Future\"},\"image\":{\"@id\":\"https:\/\/ingch.ch\/en\/#\/schema\/logo\/image\/\"},\"sameAs\":[\"https:\/\/facebook.com\/ingch2\/\",\"https:\/\/www.youtube.com\/user\/IngCH2\"]},{\"@type\":\"Person\",\"@id\":\"https:\/\/ingch.ch\/en\/#\/schema\/person\/81354b74b93673fafb73b27fb31de962\",\"name\":\"lh\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/ingch.ch\/en\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/secure.gravatar.com\/avatar\/eba92889b070ef796f1c755e80b794ce?s=96&d=mm&r=g\",\"contentUrl\":\"https:\/\/secure.gravatar.com\/avatar\/eba92889b070ef796f1c755e80b794ce?s=96&d=mm&r=g\",\"caption\":\"lh\"},\"url\":\"https:\/\/ingch.ch\/en\/author\/admin_ingch\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"IngCH","description":"Une \u00e9tude de l\u2019EPFL r\u00e9v\u00e8le pourquoi les \u00eatres humains excellent dans la reconnaissance des objets \u00e0 partir de fragments alors que l\u2019IA peine","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/","og_locale":"en_US","og_type":"article","og_title":"IngCH","og_description":"Une \u00e9tude de l\u2019EPFL r\u00e9v\u00e8le pourquoi les \u00eatres humains excellent dans la reconnaissance des objets \u00e0 partir de fragments alors que l\u2019IA peine","og_url":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/","og_site_name":"IngCH","article_publisher":"https:\/\/facebook.com\/ingch2\/","article_published_time":"2025-08-07T12:00:52+00:00","article_modified_time":"2025-08-07T13:59:30+00:00","og_image":[{"width":688,"height":386,"url":"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/688x387-1.jpg","type":"image\/jpeg"}],"author":"lh","twitter_card":"summary_large_image","twitter_misc":{"Written by":"lh","Est. reading time":"5 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#article","isPartOf":{"@id":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/"},"author":{"name":"lh","@id":"https:\/\/ingch.ch\/en\/#\/schema\/person\/81354b74b93673fafb73b27fb31de962"},"headline":"Comment faire pour que l’IA voie aussi bien que les humains","datePublished":"2025-08-07T12:00:52+00:00","dateModified":"2025-08-07T13:59:30+00:00","mainEntityOfPage":{"@id":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/"},"wordCount":887,"publisher":{"@id":"https:\/\/ingch.ch\/en\/#organization"},"image":{"@id":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#primaryimage"},"thumbnailUrl":"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/688x387-1.jpg","keywords":["Nachwuchsf\u00f6rderung","Technik","Frauen","F\u00f6rderung"],"articleSection":["IngFLASH","News","Science & Technology"],"inLanguage":"en-US"},{"@type":"WebPage","@id":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/","url":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/","name":"IngCH","isPartOf":{"@id":"https:\/\/ingch.ch\/en\/#website"},"primaryImageOfPage":{"@id":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#primaryimage"},"image":{"@id":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#primaryimage"},"thumbnailUrl":"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/688x387-1.jpg","datePublished":"2025-08-07T12:00:52+00:00","dateModified":"2025-08-07T13:59:30+00:00","description":"Une \u00e9tude de l\u2019EPFL r\u00e9v\u00e8le pourquoi les \u00eatres humains excellent dans la reconnaissance des objets \u00e0 partir de fragments alors que l\u2019IA peine","breadcrumb":{"@id":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#primaryimage","url":"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/688x387-1.jpg","contentUrl":"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/688x387-1.jpg","width":688,"height":386,"caption":"(c) EPFL\/iStock (Vertigo3d)"},{"@type":"BreadcrumbList","@id":"https:\/\/ingch.ch\/en\/ingflash\/science-technology\/comment-faire-pour-que-lia-voie-aussi-bien-que-les-humains\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Startseite","item":"https:\/\/ingch.ch\/en\/"},{"@type":"ListItem","position":2,"name":"Comment faire pour que l’IA voie aussi bien que les humains"}]},{"@type":"WebSite","@id":"https:\/\/ingch.ch\/en\/#website","url":"https:\/\/ingch.ch\/en\/","name":"IngCH","description":"MINT for our future","publisher":{"@id":"https:\/\/ingch.ch\/en\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/ingch.ch\/en\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/ingch.ch\/en\/#organization","name":"IngCH Engineers Shape our Future","url":"https:\/\/ingch.ch\/en\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/ingch.ch\/en\/#\/schema\/logo\/image\/","url":"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/2021\/06\/ingch_logo-RGB.svg","contentUrl":"https:\/\/ingch.ch\/wp\/wp-content\/uploads\/2021\/06\/ingch_logo-RGB.svg","width":"156","height":"87","caption":"IngCH Engineers Shape our Future"},"image":{"@id":"https:\/\/ingch.ch\/en\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/facebook.com\/ingch2\/","https:\/\/www.youtube.com\/user\/IngCH2"]},{"@type":"Person","@id":"https:\/\/ingch.ch\/en\/#\/schema\/person\/81354b74b93673fafb73b27fb31de962","name":"lh","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/ingch.ch\/en\/#\/schema\/person\/image\/","url":"https:\/\/secure.gravatar.com\/avatar\/eba92889b070ef796f1c755e80b794ce?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/eba92889b070ef796f1c755e80b794ce?s=96&d=mm&r=g","caption":"lh"},"url":"https:\/\/ingch.ch\/en\/author\/admin_ingch\/"}]}},"_links":{"self":[{"href":"https:\/\/ingch.ch\/en\/wp-json\/wp\/v2\/posts\/12847","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ingch.ch\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/ingch.ch\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/ingch.ch\/en\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/ingch.ch\/en\/wp-json\/wp\/v2\/comments?post=12847"}],"version-history":[{"count":3,"href":"https:\/\/ingch.ch\/en\/wp-json\/wp\/v2\/posts\/12847\/revisions"}],"predecessor-version":[{"id":12862,"href":"https:\/\/ingch.ch\/en\/wp-json\/wp\/v2\/posts\/12847\/revisions\/12862"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/ingch.ch\/en\/wp-json\/wp\/v2\/media\/12853"}],"wp:attachment":[{"href":"https:\/\/ingch.ch\/en\/wp-json\/wp\/v2\/media?parent=12847"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/ingch.ch\/en\/wp-json\/wp\/v2\/categories?post=12847"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/ingch.ch\/en\/wp-json\/wp\/v2\/tags?post=12847"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}