Objective measures for predicting speech intelligibility in noisy conditions based on new band-importance functions.

By Jianfen Ma, Y. Hu, P. Loizou. 2009

474 citations recorded by the source corpus.

Subfields: Biomedical Engineering, Cognitive Neuroscience, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

In Cognitive Neuroscience

In Signal Processing

Citations in the corpus

Time-Frequency Masking in the Complex Domain for Speech Dereverberation and Denoising (2017, neutral): “In addition, we evaluate the frequency-weighted segmental signal-to-noise ratio (SNRf w ) [25], which computes and then averages the weighted signal-to-noise ratio in each critical band.”
Deep Learning for Environmentally Robust Speech Recognition (2017, methodology): “lity assessment in place of subjective listening tests. In some evaluations, PESQ has also shown some potential in predicting speech intelligibility but this appears to depend on the noise conditions [52]. In the literature, many unsupervised signal processing methods have been developed for removin”
An Algorithm for Predicting the Intelligibility of Speech Masked by Modulated Noise Maskers (2016, background): “5 , respectively, in [39, Table IV].”
Speech Quality Assessment (2011, background): “The PESQ measure has been proven to be the most reliable objective measure for assessment of speech quality [30,34], and to some degree, speech intelligibility [74].”
Reasons why current speech-enhancement algorithms do not improve speech intelligibility and suggested solutions (2011, methodology): “To gain a better understanding on the impact of the two distortions on speech intelligibility, we use an objective function that has been found to correlate highly with speech intelligibility [9].”
Speech dereverberation and denoising using complex ratio masks (2017, methodology): “The perceptual evaluation of speech quality (PESQ) [22] and the frequency-weighted segmental signal-to-noise ratio (SNRfw) [23] are used to evaluate performance.”
Learning Spectral Mapping for Speech Dereverberation and Denoising (2015, methodology): “We quantitatively evaluate our approach by two objective measurements of speech intelligibility: frequency-weighted segmental speech-to-noise ratio ( ) [22] and short-time objective intelligibility measure (STOI) [34].”
An Algorithm for Intelligibility Prediction of Time–Frequency Weighted Noisy Speech (2011, background): “In a recent study, Ma et al. [14] showed that several intelligibility measures could benefit from the use of new (signal-dependent ) band-importance functions (BIF).”