Convolutional Deep Belief Networks for Scalable Unsupervised Learning of Hierarchical Representations

There has been much interest in unsupervised learning of hierarchical generative models such as deep belief networks. Scaling such models to full-sized, high-dimensional images remains a difficult