Related Experiment Videos
Classification of background noises for hearing-aid applications
1Center for Research in Speech and Hearing Sciences, City University of New York, Graduate Center, New York 10036.
The Journal of the Acoustical Society of America
|January 1, 1995
Summary
A new hearing aid procedure accurately classifies background noise using four signal features. This noise classification exceeds 90% accuracy, enabling adaptive hearing aid adjustments for improved listening experiences.
Area of Science:
- Acoustics and Signal Processing
- Hearing Aid Technology
- Auditory Perception
Background:
- Hearing aids require adaptive responses to changing noise environments.
- Effective background-noise classification is crucial for optimizing hearing aid performance.
Purpose of the Study:
- To develop and evaluate a background-noise classification procedure for hearing-aid applications.
- To determine the accuracy of a simplified classification method compared to a more complex one.
Main Methods:
- Measured four signal features: envelope fluctuations, mean frequency, and spectral slopes.
- Investigated envelope modulation spectra in auditory critical bands.
- Utilized cluster analysis to group feature vectors of eleven everyday background noises.
Main Results:
- The simpler four-feature classification procedure showed no disadvantage compared to the complex method.
- Noise-classification accuracy exceeded 90% when noises were grouped into reduced clusters.
- Optimal conditions involved different portions of the same noise recording for training and testing.
Conclusions:
- A robust and accurate background-noise classification procedure for hearing aids has been developed.
- The four-feature method provides a practical and effective approach for adaptive hearing aid adjustments.
- This technology promises enhanced audibility and speech intelligibility in diverse noise conditions.