Doctoral school course project. May also be part of Master thesis project 2008-2009
Navigation of large-scale image databases
General description
In browsing large image databases, it is important to be able to
locate and identify images of interest. This represents a significant
challenge since the images can be considered to be objects in a very
high-dimensional space and therefore navigating such a database is a
challenging task.
Goal
In this project we develop on recent ideas based on similarity detection using metric embeddings. This along with novel use of user input and automatic feature extraction will allow us to develop navigation algorithms.
Professor
Prof. Suhas Diggavi, LICOS, office INR 112, tel (021) 693 8301, suhas.diggavi@epfl.ch