Compact Part-Based Image Representations: Extremal Competition and Overgeneralization

Learning compact and interpretable representations for images is a very natural task which has not been solved satisfactorily even for simple classes. In this paper, we review various ways of composing parts (or experts) and argue that competitive forms of interaction are best suited to learn low-dimensional representations. Using a process of overgeneralization and correction we show in experiments that very intuitive models can be obtained.