Sequential neural models with stochastic layers

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

  • Marco Fraccaro
  • Søren Kaae Sønderby
  • Ulrich Paquet
  • Winther, Ole

How can we efficiently propagate uncertainty in a latent state representation with recurrent neural networks? This paper introduces stochastic recurrent neural networks which glue a deterministic recurrent neural network and a state space model together to form a stochastic and sequential neural generative model. The clear separation of deterministic and stochastic layers allows a structured variational inference network to track the factorization of the model's posterior distribution. By retaining both the nonlinear recursive structure of a recurrent neural network and averaging over the uncertainty in a latent path, like a state space model, we improve the state of the art results on the Blizzard and TIMIT speech modeling data sets by a large margin, while achieving comparable performances to competing methods on polyphonic music modeling.

Original languageEnglish
Title of host publicationNeural Information Processing Systems 2016
EditorsD. D. Lee, M. Sugiyama, U. V. Luxburg, I. Guyon, R. Garnett
Number of pages9
PublisherNeural Information Processing Systems Foundation
Publication date2016
Publication statusPublished - 2016
Event30th Annual Conference on Neural Information Processing Systems - Barcelona, Spain
Duration: 5 Dec 201610 Dec 2016
Conference number: 30


Conference30th Annual Conference on Neural Information Processing Systems
SeriesAdvances in Neural Information Processing Systems

ID: 179362819