Fully automated speaker identification and intelligibility assessment in dysarthria disease using auditory knowledge

Dysarthric speaker identification with different degrees of dysarthria severity using deep belief networks

“…Because dysarthric individuals are often physically incapacitated and unable to use a computer, speech processing applications, such as speech recognition and speaker recognition systems, can be helpful for dysarthric individuals. The best result achieved in this study was a speaker identification rate of 97.2% [5]. Because of this unsatisfactory performance in terms of accuracy, it is necessary to design a new speaker-recognition system for dysarthric individuals.…”

“…It is associated with unusual phonation and amplitude, especially on explosive phonemes [1]. The best result achieved in this study was a speaker identification rate of 97.2% [5]. The human auditory system may not be able to successfully identify these people by their voices because of these unnatural variations.…”

“…Dysarthria may be congenital or may be triggered by other diseases, causing unnatural variations in the voices of dysarthric individuals. To the best of our knowledge, only one significant study has been reported in the literature on this issue [5], where a hybrid Gaussian mixture model-support vector machine (GMM-SVM) system was proposed for dysarthric speaker identification. Because dysarthric individuals are often physically incapacitated and unable to use a computer, speech processing applications, such as speech recognition and speaker recognition systems, can be helpful for dysarthric individuals.…”

“…However, speaker recognition for dysarthric individuals is a challenging issue [3,4]. To the best of our knowledge, only one significant study has been reported in the literature on this issue [5], where a hybrid Gaussian mixture model-support vector machine (GMM-SVM) system was proposed for dysarthric speaker identification. In this work, the two dysarthria databases of Nemours [6] and TORGO [7] were used to train and evaluate the proposed system.…”

“…

Dysarthria is a degenerative disorder of the central nervous system that affects the control of articulation and pitch; therefore, it affects the uniqueness of sound produced by the speaker. To the best of our knowledge, only one significant study has been reported in the literature on this issue [5], where a hybrid Gaussian mixture model-support vector machine (GMM-SVM) system was proposed for dysarthric speaker identification. The effectiveness of the proposed method is demonstrated and compared with well-known Mel-frequency cepstral coefficient features.

…”