We present an approach to low-level vision that combines two main ideas: the use of convolutional networks as an image processing architecture and an unsupervised learning procedu...
In this paper, we present a novel approach to learning semantic localized patterns with binary projections in a supervised manner. The pursuit of these binary projections is refor...
Most data mining operations include an integral search component at their core. For example, the performance of similarity search or classification based on Nearest Neighbors is ...