On the compression of recurrent neural networks with an application to LVCSR acoustic modeling for embedded speech recognition

Rohit Prabhavalkar; Ouais Alsharif; Antoine Bruguier; Ian McGraw

ICASSP 2016accepted0 citations

On the compression of recurrent neural networks with an application to LVCSR acoustic modeling for embedded speech recognition

Rohit Prabhavalkar, Ouais Alsharif, Antoine Bruguier, Ian McGraw

Abstract

We study the problem of compressing recurrent neural networks (RNNs). In particular, we focus on the compression of RNN acoustic models, which are motivated by the goal of building compact and accurate speech recognition systems which can be run efficiently on mobile devices. In this work, we present a technique for general recurrent model compression that jointly compresses both recurrent and non-recurrent inter-layer weight matrices. We find that the proposed technique allows us to reduce the size of our Long Short-Term Memory (LSTM) acoustic model to a third of its original size with negligible loss in accuracy.

BibTeX

@inproceedings{icassp2016_onthecompression,
  title = {On the compression of recurrent neural networks with an application to LVCSR acoustic modeling for embedded speech recognition},
  author = {Rohit Prabhavalkar and Ouais Alsharif and Antoine Bruguier and Ian McGraw},
  booktitle = {ICASSP 2016},
  year = {2016}
}