In this chapter we review the problem of object class recognition in large image collections.We focus specifically on scenarios where the classes to be recognized are not known in advance. The motivating application is "object-class search by example" where a user provides at query time a small set of training images defining an arbitrary novel category and the system must retrieve images belonging to this class from a large database. This setting poses challenging requirements on the system design: the object classifier must be learned efficiently at query time from few examples; recognition must have low computational cost with respect to the database size; finally, compact image descriptors must be used to allow storage of large collections in memory. We review a method that addresses these requirements by learning a compact image descriptor - classemes - yielding good categorization accuracy even with efficient linear classifiers. We also study how data structures and methods from text-retrieval can be adapted to enable efficient search of an object-class in collections of several million images. © 2014 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Torresani, L., Szummer, M., & Fitzgibbon, A. (2014). Classemes: A compact image descriptor for efficient novel-class recognition and search. Studies in Computational Intelligence. Springer Verlag. https://doi.org/10.1007/978-3-642-44907-9_5
Mendeley helps you to discover research relevant for your work.