Simple spectral subtraction method enhances speech intelligibility in noise for cochlear implant listeners
It has been demonstrated that while clean speech is well intelligible by most cochlear implant (CI) listeners, noise quickly degrades speech intelligibility. To remedy the situation, CI manufacturers integrate noise reduction (NR) algorithms (often using multiple microphones) in their CI processors, and they report that CI users benefit from this measure. We have implemented a single-microphone NR scheme based on spectral subtraction with minimum statistics to see if such a simple algorithm can also effectively increase speech intelligibility in noise. We measured speech reception thresholds using both speech-shaped and car noise in 5 CI users and 23 normal-hearing listeners. For the latter group, CI hearing was acoustically simulated. In case of the CI users, the performance of the proposed NR algorithm was also compared to that of the CI processor’s built-in one. Our NR algorithm enhances intelligibility greatly in combination with the acoustic simulation regardless of the noise type; these effects are highly significant. For the CI users, trends agree with the above finding (for both the proposed and the built-in NR algorithms), however, due to low sample number, these differences did not reach statistical significance. We conclude that simple spectral subtraction can enhance speech intelligibility in noise for CI listeners and may even keep up with proprietary NR algorithms.
Boll, S. (1979). “Suppression of acoustic noise in speech using spectral subtraction,” IEEE T. Acoust. Speech, 27, 113-120.
Chilian A., Braun E., and Harczos T. (2011). “Acoustic simulation of cochlear implant hearing,” Proc. of 3rd International Symposium on Auditory and Audiological Research (ISAAR 2011) – Speech Perception and Auditory Disorders, Nyborg, Denmark, pp. 425-432.
Liu, D., Smaragdis P., and Kim M. (2014). “Experiments on deep learning for speech denoising,” Proc Annual Conference of the International Speech Communication Association (INTERSPEECH), Singapore.
Loizou, P.C. (2007). Speech Enhancement: Theory and Practice. CRC Press: Boca Raton, FL, USA.
Martin, R. (1994). “Spectral subtraction based on minimum statistics,” Proc. European Signal Processing Conference (EUSIPCO) ’94, Edinburgh, Scotland, nr. 1, pp. 1182-1185.
Wagener, K., Kühnel, V., and Kollmeier, B. (1999). “Entwicklung und Evaluation eines Satztests in deutscher Sprache I: Design des Oldenburger Satztests (Development and evaluation of a sentence test in German language I: Design of the Oldenburg sentence test),” Z. Audiol, 38, 4-15.
Zoghlami, N., Lachiri Z., and Ellouze N. (2010). “Perceptually motivated generalized spectral subtraction for speech enhancement,” Advances in Nonlinear Speech Processing, Lecture Notes Artif. Int., 5933, 136-143.
Authors who publish with this journal agree to the following terms:
a. Authors retain copyright* and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
b. Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
c. Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
*From the 2017 issue onward. The Danavox Jubilee Foundation owns the copyright of all articles published in the 1969-2015 issues. However, authors are still allowed to share the work with an acknowledgement of the work's authorship and initial publication in this journal.