Gaston Hilkhuysen

Learn More
We propose a data driven, non-intrusive method for speech intelligibility estimation. We begin with a large set of speech signal specific features and use a dimensionality reduction approach based on correlation and principal component analysis to find the most relevant features for intelligibility prediction. These are then used to train a Gaussian mixture(More)
The effects on speech intelligibility of three different noise reduction algorithms (spectral subtraction, minimal mean squared error spectral estimation, and subspace analysis) were evaluated in two types of noise (car and babble) over a 12 dB range of signal-to-noise ratios (SNRs). Results from these listening experiments showed that most algorithms(More)
Using the data presented in the accompanying paper [Hilkhuysen et al., J. Acoust. Soc. Am. 131, 531-539 (2012)], the ability of six metrics to predict intelligibility of speech in noise before and after noise suppression was studied. The metrics considered were the Speech Intelligibility Index (SII), the fractional Articulation Index (fAI), the coherence(More)
Existing measures of speech intelligibility and speech quality can be ineffective for evaluating new types of speech communication systems, such as wideband audio codecs, digital hearing aids and noise-reduction systems. We propose that new performance-based evaluation methods are required which tap into the cognitive effort listeners employ to understand(More)
This paper presents the development of a compact vocabulary for describing the audible characteristics of degraded speech. An experiment was conducted with 51 English-speaking subjects who were tasked with assigning one of a list of given text descriptors to 220 degradation conditions. Exploratory data analysis using hierarchical clustering resulted in a(More)
The effect of noise reduction on the intelligibility of speech in noise is poorly understood. Although the SNR of noisy speech is improved by the removal of more noise than speech from the signal, the expected increase in intelligibility does not typically occur. To account for these deleterious effects we present an orthogonal decomposition of the signal(More)
PURPOSE In this study, the authors investigated how well experts can adjust the settings of a commercial noise-reduction system to optimize the intelligibility for naive normal-hearing listeners. METHOD In Experiment 1, 5 experts adjusted parameters for a noise-reduction system while aiming to optimize intelligibility. The stimuli consisted of speech(More)
  • 1