RT Journal Article
T1 An attention Long Short-Term Memory based system for automatic classification of speech intelligibility
A1 Fernandez Diaz, Miguel
A1 Gallardo Antolín, Ascensión
AB Speech intelligibility can be degraded due to multiple factors, such as noisy environments, technical difficulties or biological conditions. This work is focused on the development of an automatic non-intrusive system for predicting the speech intelligibility level in this latter case. The main contribution of our research on this topic is the use of Long Short-Term Memory (LSTM) networks with log-mel spectrograms as input features for this purpose. In addition, this LSTM-based system is further enhanced by the incorporation of a simple attention mechanism that is able to determine the more relevant frames to this task. The proposed models are evaluated with the UA-Speech database that contains dysarthric speech with different degrees of severity. Results show that the attention LSTM architecture outperforms both, a reference Support Vector Machine (SVM)-based system with hand-crafted features and a LSTM-based system with Mean-Pooling.
PB Elsevier
SN 0952-1976
YR 2020
FD 2020-11
LK https://hdl.handle.net/10016/33236
UL https://hdl.handle.net/10016/33236
LA eng
NO The work leading to these results has been partly supported by the Spanish Government-MinECo under Project TEC2017-84395-P. The authors wish to acknowledge Dr. Mark Hasegawa-Johnson for making the UA-Speech database available.
DS e-Archivo
RD 27 jul. 2024