Detail of the student project

Topic:[Deep Learning] Semantic Hashing
Department:Katedra kybernetiky
Supervisor:Mgr. Oleksandr Shekhovtsov, Ph.D.
Announce as:Diplomová práce, Bakalářská práce, Semestrální projekt
Description:We will train a neural network to produce binary hash codes. Such codes can be used for a very efficient look-up, much like indexing in a (hierarchical) catalogue, of semantically similar items (documents, images, image patches). You will learn gradient estimators of binary variables, design models that cluster data into (hierarchical) categories, experiment with new models and learning methods, with space to propose own practical and theoretical improvements.
Responsible person: Petr Pošík