Evaluation of formant-like features on an automatic vowel classification task.

Numerous attempts have been made to find low-dimensional, formant-related representations of speech signals that are suitable for automatic speech recognition. However, it is often not known how these features behave in comparison with true formants. The purpose of this study was to compare two sets of automatically extracted formant-like features, i.e., robust formants and HMM2 features, to hand-labeled formants. The robust formant features were derived by means of the split Levinson algorithm while the HMM2 features correspond to the frequency segmentation of speech signals obtained by two-dimensional hidden Markov models. Mel-frequency cepstral coefficients (MFCCs) were also included in the investigation as an example of state-of-the-art automatic speech recognition features. The feature sets were compared in terms of their performance on a vowel classification task. The speech data and hand-labeled formants that were used in this study are a subset of the American English vowels database presented in Hillenbrand et al. [J. Acoust. Soc. Am. 97, 3099-3111 (1995)]. Classification performance was measured on the original, clean data and in noisy acoustic conditions. When using clean data, the classification performance of the formant-like features compared very well to the performance of the hand-labeled formants in a gender-dependent experiment, but was inferior to the hand-labeled formants in a gender-independent experiment. The results that were obtained in noisy acoustic conditions indicated that the formant-like features used in this study are not inherently noise robust. For clean and noisy data as well as for the gender-dependent and gender-independent experiments the MFCCs achieved the same or superior results as the formant features, but at the price of a much higher feature dimensionality.

Some objective measures indicative of perceived voice robustness in student teachers.

One of the problems confronted in the teaching profession is the maintenance of a healthy voice. This basic pedagogical tool is subjected to extensive use, and frequently suffers from overload, with some teachers having to give up their profession altogether. In some teacher training schools, it is the current practice to examine the student's voice, and to refer any perceived susceptibility to strain to voice specialists. For this study, a group of vocally healthy students were examined first at the teacher training schools, and then at the ENT clinic at the University Hospital of Nijmegen. The aim was to predict whether the subject's voice might be at risk for occupational dysphonia as a result of the vocal load of the teaching profession. We tried to find objective measures of voice quality in student teachers, used in current clinical practice, which reflect the judgements of the therapists and phoniatricians. We tried to explain such measures physiologically in terms of robustness of, and control over voicing. Objective measures used included video-laryngostroboscopy, phonetography and spectrography. Maximum phonation time, melodic range in conjunction with maximum intensity range, and the production of soft voice are suggested as possible predictive parameters for the risk of occupational voice strain.