Predicting the quality of enhanced wideband speech with a cochlear model Journal Articles

abstract

• Objective measures are commonly used in the development of speech coding algorithms as an adjunct to human subjective evaluation. Predictors of speech quality based on models of physiological or perceptual processing tend to perform better than measures based on simple acoustical properties. Here, a modeling method based on a detailed physiological model and a neurogram similarity measure is developed and optimized to predict the quality of an enhanced wideband speech dataset. A model capturing temporal modulations in neural activity up to 267 Hz was found to perform as well as or better than several existing objective quality measures.

authors

• Wirtzfeld, Michael R
• Pourmand, Nazanin
• Parsa, Vijay
• Bruce, Ian

status

• published 

publication date

• September 1, 2017

has subject area

• Acoustics (Science Metrix)
• Algorithms (MeSH)
• Auditory Perception (MeSH)
• Cochlea (MeSH)
• Cochlear Nerve (MeSH)
• Datasets as Topic (MeSH)
• Female (MeSH)
• Hearing (MeSH)
• Humans (MeSH)
• Linear Models (MeSH)
• Male (MeSH)
• Models, Biological (MeSH)
• Noise (MeSH)
• Perceptual Masking (MeSH)
• Signal-To-Noise Ratio (MeSH)
• Speech Intelligibility (MeSH)

published in

• Journal of the Acoustical Society of America  Journal

keywords

• Algorithms
• Auditory Perception
• Cochlea
• Cochlear Nerve
• Datasets as Topic
• Female
• Hearing
• Humans
• Linear Models
• Male
• Models, Biological
• Noise
• Perceptual Masking
• Signal-To-Noise Ratio
• Speech Intelligibility

Digital Object Identifier (DOI)

• 10.1121/1.5003785

PubMed ID

• 28964067

start page

• EL319

end page

• EL325

volume

• 142

issue

• 3