Investigating gated recurrent networks for speech synthesis

Recently, recurrent neural networks (RNNs) as powerful sequence models have re-emerged as a potential acoustic model for statistical parametric speech synthesis (SPSS). The long short-term memory (LSTM) architecture is particularly attractive because it addresses the vanishing gradient problem in st...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhizheng Wu, King, Simon
Format: Conference Proceeding
Language:English
Subjects:
acoustic modelling
Acoustics
Architecture
Computer architecture
Electronics
Feedforward
gated recurrent network
Gates
Hidden Markov models
Logic gates
long short-term memory
Microprocessors
Neural networks
Proposals
recurrent network network
Speech
Speech recognition
Speech synthesis
Trajectory
Citations: Items that cite this one
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Recently, recurrent neural networks (RNNs) as powerful sequence models have re-emerged as a potential acoustic model for statistical parametric speech synthesis (SPSS). The long short-term memory (LSTM) architecture is particularly attractive because it addresses the vanishing gradient problem in standard RNNs, making them easier to train. Although recent studies have demonstrated that LSTMs can achieve significantly better performance on SPSS than deep feedforward neural networks, little is known about why. Here we attempt to answer two questions: a) why do LSTMs work well as a sequence model for SPSS; b) which component (e.g., input gate, output gate, forget gate) is most important. We present a visual analysis alongside a series of experiments, resulting in a proposal for a simplified architecture. The simplified architecture has significantly fewer parameters than an LSTM, thus reducing generation complexity considerably without degrading quality.
ISSN:2379-190X
DOI:10.1109/ICASSP.2016.7472657