Extracting and Composing Robust Features with DenoisingAutoencodersPascal Vincentvincentp@iro.umontreal.caHugo Larochellelarocheh@iro.umontreal.caYoshua Bengiobengioy@iro.umontreal.caPierre-Antoine Manzagolmanzagop@iro.umontreal.caUniversit´e de Montr´eal, Dept. IRO, CP 6128, Succ. Centre-Ville, Montral, Qubec, H3C 3J7, CanadaAbstractPrevious work has shown that the difficul-ties in learning deep generative or discrim-inative models can be overcome by an ini-tial unsupervised learning step that maps in-puts to useful intermediate representations.We introduceand motivate a new trainingprinciple for unsupervised learning of a rep-resentation based on the idea of making thelearned representations robust to partial cor-ruption of the input pattern. This approachcan be used to train autoencoders, and thesedenoising autoencoders can be stacked to ini-tialize deep architectures. The algorithm canbe motivated from a manifold learning andinformation theoretic perspective or from agenerative model perspective. Comparativeexperiments clearly show the surprising ad-vantage of corrupting the input of autoen-coders on a pattern classification benchmarksuite.1. IntroductionRecent theoretical studies indicate that deep architec-tures (Bengio & Le Cun, 2007; Bengio, 2007) may beneeded to effi cientlymodel complex distributions andachieve better generalization performance on challeng-ing recognition tasks. The belief that additional levelsof functional composition will yield increased repre-sentational and modeling power is not new (McClel-land et al., 1986; Hinton, 1989; Utgoff & Stracuzzi,2002). However, in practice, learning in deep archi-tectures has proven to be difficult. One needs onlyAppearing in Proceedings of the 25 th International Confer-ence on Machine Learnin...