Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
This lecture explores different types of image systems, including concept-based, named entity-based, and perspective connections-based systems. It delves into the centrality of document-images, coherent subsets in corpora, and the analysis of graphics. The lecture discusses the nature of concept-based image systems, their hierarchical structure, and their role in artificial vision. It also covers named entity-based systems, their unique identification of entities, and integration with international databases like Wikidata. Moreover, the lecture explains perspective connections-based systems, focusing on establishing visual relationships between images from different perspectives. The lecture concludes with a discussion on morphological relationships in image systems, emphasizing the formalization of morphological patterns and the challenges in detecting invariance across diverse materials and textures.