• Corpus ID: 26127229

On speech intelligibility estimation of phase-aware single-channel speech enhancement

  title={On speech intelligibility estimation of phase-aware single-channel speech enhancement},
  author={Andreas Gaich and Pejman Mowlaee Begzade Mahale},
To reduce time and costs in the development process of noise reduction algorithms, an objective intelligibility measure is crucial. Such a measure has to show high correlation with speech intelligibility determined by real listening experiments. In the past several measures were found that perform reliable in a particular scenario when only the spectral amplitude of a noisy signal is modified. Recent studies demonstrate the positive impact of a phase modification in a single-channel speech… 

Figures from this paper

Advances in phase-aware signal processing in speech communication

On the importance of harmonic phase modification for improved speech signal reconstruction

The results show that enhancement of decomposed phase parts suffices for improved reconstruction in speech enhancement and compares the proposed harmonic phase modification with other phase estimation methods.

Speech enhancement using harmonic-structure-based phase reconstruction

This paper reviews two harmonic-structure-based phase estimation methods with temporal and frequency constraints on the harmonic speech phase and describes important parameters for phase estimation, such as the frame shift length and window function of the short-time Fourier transform.

Phase Continuity: Learning Derivatives of Phase Spectrum for Speech Enhancement

An effective phase reconstruction strategy for neural speech enhancement that can operate in noisy environments is proposed and a phase continuity loss that considers relative phase variations across the time and frequency axes is introduced.

Multiframe maximum a posteriori estimators for single-microphone speech enhancement

Iran National Science Foundation (INSF) for supporting this work under, Grant/Award Number: 96000455 Abstract Multiframe maximum a posteriori (MAP) estimators are applied to a single‐microphone noise

The Conversation: Deep Audio-Visual Speech Enhancement

A deep audio-visual speech enhancement network that is able to separate a speaker's voice given lip regions in the corresponding video, by predicting both the magnitude and the phase of the target signal.



On speech quality estimation of phase-aware single-channel speech enhancement

To approximate the speech quality of a given speech enhancement system, most of the existing instrumental metrics rely on the calculation of a distortion metric defined between the clean reference

Objective Intelligibility Measures Based on Mutual Information for Speech Subjected to Speech Enhancement Processing

Using data from three different listening tests it is shown that the proposed objective intelligibility measures provide promising results for speech intelligibility prediction in different scenarios of speech enhancement where speech is processed by non-linear modification strategies.

An Algorithm for Intelligibility Prediction of Time–Frequency Weighted Noisy Speech

A short-time objective intelligibility measure (STOI) is presented, which shows high correlation with the intelligibility of noisy and time-frequency weighted noisy speech (e.g., resulting from noise reduction) of three different listening experiments and showed better correlation with speech intelligibility compared to five other reference objective intelligible models.

Phase Estimation in Single-Channel Speech Enhancement: Limits-Potential

It is demonstrated that the proposed time-frequency phase smoothing method successfully reduces the variance of the noisy phase at harmonics and balances a tradeoff between a joint improvement in perceived quality and speech intelligibility by phase-only enhancement.

Harmonic Phase Estimation in Single-Channel Speech Enhancement Using Phase Decomposition and SNR Information

By enhancing the noisy phase both perceived speech quality as well as speech intelligibility are improved as predicted by the instrumental metrics and justified by subjective listening tests.

Enhancement of speech intelligibility in near-end noise conditions with phase modification

Results indicate that the proposed post-processing method improves speech intelligibility over the reference method.

Reasons why Current Speech-Enhancement Algorithms do not Improve Speech Intelligibility and Suggested Solutions

  • P. LoizouGibak Kim
  • Physics, Computer Science
    IEEE Transactions on Audio, Speech, and Language Processing
  • 2011
A theoretical framework is presented that can be used to analyze potential factors that can influence the intelligibility of processed speech and focuses on the fine-grain analysis of the distortions introduced by speech enhancement algorithms.

Speech Intelligibility Prediction Based on Mutual Information

  • J. JensenC. Taal
  • Physics
    IEEE/ACM Transactions on Audio, Speech, and Language Processing
  • 2014
The proposed model predicts that speech intelligibility cannot be improved by any processing of noisy critical-band amplitudes, and performs well in predicting the intelligibility of speech signals contaminated by additive noise and potentially non-linearly processed using time-frequency weighting.