A hierarchy of recurrent networks for speech recognition

@inproceedings{Schrauwen2009AHO,
  title={A hierarchy of recurrent networks for speech recognition},
  author={Benjamin Schrauwen},
  booktitle={NIPS 2009},
  year={2009}
}
Generative models for sequential data based on directed gra phs of Restricted Boltzmann Machines (RBMs) are able to accurately model high d mensional sequences as recently shown. In these models, temporal depend ncies in the input are discovered by either buffering previous visible variab les or by recurrent connections of the hidden variables. Here we propose a modificat ion of these models, the Temporal Reservoir Machine (TRM). It utilizes a recurre nt artificial neural network (ANN) for… CONTINUE READING

Citations

Publications citing this paper.

Similar Papers

Loading similar papers…