{"id":84,"date":"2026-05-28T16:42:43","date_gmt":"2026-05-28T14:42:43","guid":{"rendered":"https:\/\/project.inria.fr\/ellisgrenoble\/?page_id=84"},"modified":"2026-05-28T17:29:16","modified_gmt":"2026-05-28T15:29:16","slug":"research","status":"publish","type":"page","link":"https:\/\/project.inria.fr\/ellisgrenoble\/research\/","title":{"rendered":"Research"},"content":{"rendered":"\n\n<p class=\"wp-block-paragraph\">The ELLIS Unit Grenoble&#8217;s research spans three interlocking themes, united by a shared commitment to <strong>multimodal learning<\/strong> \u2014 the science of building AI systems that learn from and reason over multiple, complementary types of data.<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\">Core Algorithmic Foundations<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">We develop the theoretical and methodological tools that underpin robust and reliable AI systems. This includes work in <strong>machine learning<\/strong>, <strong>computer vision<\/strong>, <strong>representation learning<\/strong>, <strong>domain adaptation<\/strong>, <strong>robustness<\/strong>, and <strong>interpretability<\/strong>. Our goal is to advance the science of learning itself \u2014 from modelling uncertainty and multi-view data to developing principled approaches that generalise across settings.<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\">Modalities and Their Integration<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">A defining feature of the Unit is its focus on cross-modal learning: how AI systems can integrate, align, and reason jointly over heterogeneous sources of information. Our work spans <strong>visual data<\/strong> (images, video, 3D), <strong>audio and speech<\/strong>, <strong>natural language<\/strong>, and <strong>spatial and geophysical signals<\/strong>. Understanding how these modalities complement and constrain each other is central to our scientific identity and distinguishes us within the ELLIS network.<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\">Application Domains<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Our foundational research is driven by and validated in real-world application domains of high societal relevance:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Remote sensing and computational geoscience<\/strong> \u2014 Earth observation, environmental monitoring, climate science, and hyperspectral imaging.<\/li>\n<li><strong>Robotics<\/strong> \u2014 Multimodal perception, human-robot interaction, and learning for autonomous systems.<\/li>\n<li><strong>Medical imaging<\/strong> \u2014 Diagnostic support, computational anatomy, and AI-assisted healthcare.<\/li>\n<\/ul>\n\n\n\n<p class=\"wp-block-paragraph\">We emphasise both theoretical depth and applied, real-world impact: from modelling uncertainty and multi-view temporal data to integrating multimodal signals in environments such as healthcare, environment monitoring, and remote sensing.<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\">ELLIS Programs<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Members of ELLIS Unit Grenoble participate in several ELLIS Programs, including <strong>Multimodal Learning Systems<\/strong>, <strong>Machine Learning and Computer Vision<\/strong>, <strong>Learning for Graphics and Vision<\/strong>, <strong>Machine Learning for Earth and Climate Sciences<\/strong>, and <strong>Theory, Algorithms and Computations of Modern Learning Systems<\/strong>.<\/p>\n\n","protected":false},"excerpt":{"rendered":"<p>The ELLIS Unit Grenoble&#8217;s research spans three interlocking themes, united by a shared commitment to multimodal learning \u2014 the science of building AI systems that learn from and reason over multiple, complementary types of data. Core Algorithmic Foundations We develop the theoretical and methodological tools that underpin robust and reliable\u2026<\/p>\n<p> <a class=\"continue-reading-link\" href=\"https:\/\/project.inria.fr\/ellisgrenoble\/research\/\"><span>Continue reading<\/span><i class=\"crycon-right-dir\"><\/i><\/a> <\/p>\n","protected":false},"author":1210,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":"","_members_access_role":[],"_members_access_error":""},"class_list":["post-84","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/project.inria.fr\/ellisgrenoble\/wp-json\/wp\/v2\/pages\/84","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/project.inria.fr\/ellisgrenoble\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/project.inria.fr\/ellisgrenoble\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/project.inria.fr\/ellisgrenoble\/wp-json\/wp\/v2\/users\/1210"}],"replies":[{"embeddable":true,"href":"https:\/\/project.inria.fr\/ellisgrenoble\/wp-json\/wp\/v2\/comments?post=84"}],"version-history":[{"count":2,"href":"https:\/\/project.inria.fr\/ellisgrenoble\/wp-json\/wp\/v2\/pages\/84\/revisions"}],"predecessor-version":[{"id":109,"href":"https:\/\/project.inria.fr\/ellisgrenoble\/wp-json\/wp\/v2\/pages\/84\/revisions\/109"}],"wp:attachment":[{"href":"https:\/\/project.inria.fr\/ellisgrenoble\/wp-json\/wp\/v2\/media?parent=84"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}