Learning Convolutional Feature Hierarchies for Visual Recognition

We propose an unsupervised method for learning multi-stage hierarchies of sparse convolutional features. While sparse coding has become an increasingly popular method for learning visual features, it