Predicting Preference Judgments of Individual Normal and Hearing-Impaired Listeners With Gaussian Processes

• Published in: IEEE transactions on audio, speech, and language processing Vol. 19; no. 4; pp. 811 - 821
• Main Authors: Groot, P C, Heskes, T, Dijkstra, T M H, Kates, J M
• Format: Journal Article
• Language: English
• Published: Piscataway, NJ IEEE 01.05.2011; Institute of Electrical and Electronics Engineers
• Subjects: Additive noise; Applied sciences; Auditory system; Bayes procedures; Clipping; Degradation; Distortion; Distortion measurement; Exact sciences and technology; Gaussian process (GP); Gaussian processes; Information, signal and communications theory; Judgments; Kernel; Kernels; Nonlinear distortion; Nonlinearity; pairwise comparisons; Predictive models; Probability theory; Psychoacoustic models; Signal and communications theory; Signal processing algorithms; Signal representation. Spectral analysis; Signal, noise; Sound; subjective quality measures; Telecommunications and information theory; Additive noise; Performance evaluation; Gaussian process (GP); Probabilistic approach; Acoustic signal; subjective quality measures; Bayes procedures; Sound quality; Non linear phenomenon; Auditive prosthesis; Subjective evaluation; Kernel method; Learning; Gaussian process; pairwise comparisons; Quality control; Clipping; Non linear effect; Signal analysis; Signal distortion
• ISSN: 1558-7916; 1558-7924
• DOI: 10.1109/TASL.2010.2064311

A probabilistic kernel approach to pairwise preference learning based on Gaussian processes is applied to predict preference judgments for sound quality degradation mechanisms that might be present in a hearing aid. Subjective sound quality comparisons for 14 normal-hearing and 18 hearing-impaired subjects were used for evaluating the predictive performance. Stimuli were sentences subjected to three kinds of distortion (additive noise, peak clipping, and center clipping) with eight levels of degradation for each distortion type. The kernel approach gives a significant improvement in preference predictions of hearing-impaired subjects by individualizing the learning process. A significant difference is shown between normal-hearing and hearing-impaired subjects, because of nonlinearities in the perception of hearing-impaired subjects. In particular, hearing-impaired subjects have significant nonlinear preference judgments when making pairwise comparisons between peak clipped sentences with different clipping thresholds. The probabilistic kernel approach is shown to be robust when generalizing over distortions and over subjects.