• Publications
  • Influence
Sparse representations for automatic target classification in SAR images
We propose a sparse representation approach for classifying different targets in Synthetic Aperture Radar (SAR) images. Unlike the other feature based approaches, the proposed method does not requireExpand
Automatic assessment of vowel space area.
An automated algorithm is described using healthy, connected speech rather than single syllables and estimates the entire vowel working space rather than corner vowels, revealing a strong correlation between the traditional VSA and automated estimates. Expand
Empirical Non-Parametric Estimation of the Fisher Information
This work obtains an empirical estimator of the FIM that does not require density estimation and is asymptotically consistent and empirically evaluates the validity of the approach using two experiments. Expand
Tracking discourse complexity preceding Alzheimer's disease diagnosis: a case study comparing the press conferences of Presidents Ronald Reagan and George Herbert Walker Bush.
A method to extract key features from discourse transcripts from President Ronald Reagan and President George Herbert Walker Bush, who have no known diagnosis of Alzheimer's disease, were described and regression analyses were conducted. Expand
Empirically Estimable Classification Bounds Based on a Nonparametric Divergence Measure
It is shown that a nonparametric f-divergence measure can be used to provide improved bounds on the minimum binary classification probability of error for the case when the training and test data are drawn from the same distribution. Expand
Convex Weighting Criteria for Speaking Rate Estimation
This paper proposes two convex cost functions for learning the weighting functions and an adaptation strategy to customize the approach to a particular speaker using minimal training, and shows that the proposed methods outperform three competing methods on both healthy and dysarthric speech. Expand
Experiments on Auditory-Visual Perception of Sentences by Users of Unilateral, Bimodal, and Bilateral Cochlear Implants.
The data are consistent with the view that V information improves cochlear implants users' ability to identify syllables in the acoustic stream and to recognize their relative juxtaposed strengths. Expand
Differences in fibertract profiles between patients with migraine and those with persistent post-traumatic headache
Although for both migraine and persistent post-traumatic headache there was a positive relationship between Fibertract alterations and headache frequency, there were disease-specific differences between headache frequency and fibertract injury patterns. Expand
Acoustic and perceptual speech characteristics of native Mandarin speakers with Parkinson's disease.
Speakers with PD exhibited significant differences in fundamental frequency, pitch variation, vowel space, and rate relative to controls, however, in contrast to the English studies, speech rate was consistently slow and most strongly correlated with intelligibility. Expand
Online speaking rate estimation using recurrent neural networks
An evaluation on spontaneous speech shows that the proposed online speaking rate estimation model based on recurrent neural networks yields a higher correlation between the estimated rate and the ground-truth rate when compared to the state-of-the-art alternatives. Expand