Hybrid Speech Enhancement with Wiener filters and Deep LSTM Denoising Autoencoders

Coto-Jiménez, Marvin; Goddard-Close, John; Persia, Leandro E. Di; Rufiner, Hugo Leonardo

doi:10.1109/iwobi.2018.8464132

Cited by 15 publications

(7 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Two layers of a bi-directional Long Short-Term Memory (bi-LSTM) with 64 filters and 32 filters are included in the encoder network to reconstruct a clean MFCC representation, followed by a temporal Fully Connected (FC) in a decoder network. An LSTM neural network was previously used to extract speaker representation and remove noise from audio [30] – [32] . In this study, Deep-MASKS employs bi-LSTM, a variation of LSTM that considers the backward sequence in addition to the forward sequence, to learn temporal patterns over the MFCC sequence.…”

Section: Methodsmentioning

confidence: 99%

Privacy-Preserving Deep Speaker Separation for Smartphone-Based Passive Speech Assessment

Ditthapron

Agu

Lammert

2021

IEEE Open J. Eng. Med. Biol.

View full text Add to dashboard Cite

Goal: Smartphones can be used to passively assess and monitor patients’ speech impairments caused by ailments such as Parkinson’s disease, Traumatic Brain Injury (TBI), Post-Traumatic Stress Disorder (PTSD) and neurodegenerative diseases such as Alzheimer’s disease and dementia. However, passive audio recordings in natural settings often capture the speech of non-target speakers (cross-talk). Consequently, speaker separation, which identifies the target speakers’ speech in audio recordings with two or more speakers’ voices, is a crucial pre-processing step in such scenarios. Prior speech separation methods analyzed raw audio. However, in order to preserve speaker privacy, passively recorded smartphone audio and machine learning-based speech assessment are often performed on derived speech features such as Mel-Frequency Cepstral Coefficients (MFCCs). In this paper, we propose a novel Deep MFCC bAsed SpeaKer Separation (Deep-MASKS). Methods: Deep-MASKS uses an autoencoder to reconstruct MFCC components of an individual’s speech from an i-vector, x-vector or d-vector representation of their speech learned during the enrollment period. Deep-MASKS utilizes a Deep Neural Network (DNN) for MFCC signal reconstructions, which yields a more accurate, higher-order function compared to prior work that utilized a mask. Unlike prior work that operates on utterances, Deep-MASKS operates on continuous audio recordings. Results: Deep-MASKS outperforms baselines, reducing the Mean Squared Error (MSE) of MFCC reconstruction by up to 44% and the number of additional bits required to represent clean speech entropy by 36%.

show abstract

Section: Methodsmentioning

confidence: 99%

Privacy-Preserving Deep Speaker Separation for Smartphone-Based Passive Speech Assessment

Ditthapron

Agu

Lammert

2021

IEEE Open J. Eng. Med. Biol.

View full text Add to dashboard Cite

show abstract

“…An autoencoder for noise reduction is a neural network architecture that has been successful in various tasks related to speech [29]. This architecture consists of an encoder that transforms an input vector s into a representation in the hidden layers h through a f mapping.…”

Section: Autoencoders Of Blstm Networkmentioning

confidence: 99%

Evaluation of Mixed Deep Neural Networks for Reverberant Speech Enhancement

Gutiérrez-Muñoz¹,

González-Salazar²,

Coto-Jiménez³

2019

Preprint

View full text Add to dashboard Cite

Speech signals are degraded in real-life environments, product of background noise or other factors. The processing of such signals for voice recognition and voice analysis systems presents important challenges. One of the conditions that make adverse quality difficult to handle in those systems is reverberation, produced by sound wave reflections that travel from the source to the microphone in multiple directions.To enhance signals in such adverse conditions, several deep learning-based methods have been proposed and proven to be effective. Recently, recurrent neural networks, especially those with long and short-term memory (LSTM), have presented surprising results in tasks related to time-dependent processing of signals, such as speech. One of the most challenging aspects of LSTM networks is the high computational cost of the training procedure, which has limited extended experimentation in several cases. In this work, we present a proposal to evaluate the hybrid models of neural networks to learn different reverberation conditions without any previous information. The results show that some combination of LSTM and perceptron layers produce good results in comparison to those from pure LSTM networks, given a fixed number of layers. The evaluation has been made based on quality measurements of the signal's spectrum, training time of the networks and statistical validation of results. Results help to affirm the fact that hybrid networks represent an important solution for speech signal enhancement, with advantages in efficiency, but without a significan drop in quality.

show abstract

“…One of the main architectures applied for regression tasks (including speech enhancement) using deep neural networks are the autoencoders. An autoencoder for speech enhancement is a neural network architecture that has been successful in various tasks related to speech [33]. This architecture consists of an encoder that transforms an input vector s into a representation in the hidden layers h through a f mapping.…”

Section: Autoencoders Of Blstm Networkmentioning

confidence: 99%

Evaluation of Mixed Deep Neural Networks for Reverberant Speech Enhancement

2019

View full text Add to dashboard Cite

Speech signals are degraded in real-life environments, as a product of background noise or other factors. The processing of such signals for voice recognition and voice analysis systems presents important challenges. One of the conditions that make adverse quality difficult to handle in those systems is reverberation, produced by sound wave reflections that travel from the source to the microphone in multiple directions. To enhance signals in such adverse conditions, several deep learning-based methods have been proposed and proven to be effective. Recently, recurrent neural networks, especially those with long short-term memory (LSTM), have presented surprising results in tasks related to time-dependent processing of signals, such as speech. One of the most challenging aspects of LSTM networks is the high computational cost of the training procedure, which has limited extended experimentation in several cases. In this work, we present a proposal to evaluate the hybrid models of neural networks to learn different reverberation conditions without any previous information. The results show that some combinations of LSTM and perceptron layers produce good results in comparison to those from pure LSTM networks, given a fixed number of layers. The evaluation was made based on quality measurements of the signal’s spectrum, the training time of the networks, and statistical validation of results. In total, 120 artificial neural networks of eight different types were trained and compared. The results help to affirm the fact that hybrid networks represent an important solution for speech signal enhancement, given that reduction in training time is on the order of 30%, in processes that can normally take several days or weeks, depending on the amount of data. The results also present advantages in efficiency, but without a significant drop in quality.

show abstract

Hybrid Speech Enhancement with Wiener filters and Deep LSTM Denoising Autoencoders

Cited by 15 publications

References 11 publications

Privacy-Preserving Deep Speaker Separation for Smartphone-Based Passive Speech Assessment

Privacy-Preserving Deep Speaker Separation for Smartphone-Based Passive Speech Assessment

Evaluation of Mixed Deep Neural Networks for Reverberant Speech Enhancement

Evaluation of Mixed Deep Neural Networks for Reverberant Speech Enhancement

Contact Info

Product

Resources

About