NMF-based Improvement of DNN and LSTM Pre-Training for Speech Enhancemet

NMF-based Improvement of DNN and LSTM Pre-Training for Speech Enhancemet

A novel pre-training method is proposed to improve deep-neural-networks (DNN) and long-short-term-memory (LSTM) performance, and reduce the local minimum problem for speech enhancement. We propose initializing the last layer weights of DNN and LSTM by Non-Negative-Matrix-Factorization (NMF) basis tr...

Full description

Bibliographic Details
Main Authors: Razieh Safari Dehnavi, Sanaz Seyedin
Format: Article
Language:English
Published: Iran Telecom Research Center 2023-09-01
Series:International Journal of Information and Communication Technology Research
Subjects:
pre-training
deep neural networks (dnn)
long short-term memory (lstm)
non-negative matrix factorization (nmf)
speech enhancement
basis matrix
noise classification
Online Access:http://ijict.itrc.ac.ir/article-1-555-en.pdf

Internet

http://ijict.itrc.ac.ir/article-1-555-en.pdf

Similar Items