About
Dysarthria is a motor speech disorder which occurs when the muscles used for speaking weaken or the person has some difficulty in controlling and coordinating the muscles. Neurodegenerative diseases, like Amyotrophic lateral sclerosis (ALS) and Parkinson’s disease (PD), can cause dysarthria. This disorder affects various aspects of speech production like articulation, phonation, prosody, respiration and resonance. As a result, intelligibility, naturalness and often audibility of speech get compromised, which negatively impact the efficiency of vocal communication for these patients.
The figure below shows the waveforms and associated spectrograms of speech obtained from a dysarthric ALS patient, a dysarthric PD patient and a healthy individual. Each person is uttering the syllable ‘pa’ in succession. It can be observed that the speaking rate significantly reduces for the patients and the information content in the spectrogram becomes confined to lower frequency regions.
At SPIRE Lab, we aim to understand such characteristics of dysarthric speech from the perspectives of signal processing as well as speech production. We further aim to utilize these insights in developing speech-based advanced tools and techniques which can enhance the quality of life of the individuals suffering from this condition.
Objectives:
The objectives of our research include, but are not limited to, the following.
- To understand and model the effects of dysarthria and the related neurodegenerative disorders on different parts of the speech production system
- To develop resource efficient algorithms for speech-based automatic detection and monitoring of the neurodegenerative disorders
- To develop data-driven methods of enhancing the naturalness and intelligibility of dysarthric speech
- To design automatic speech recognition systems for this type of disordered speech
Background and Motivation:
Till date, there exists no particular laboratory test that can confirm neurodegenerative diseases like ALS and PD. The clinical diagnosis and monitoring processes are highly time-consuming and prone to biases due to subjective perceptions. Thus, automated and efficient diagnostic tools for these diseases are the need of the hour. Moreover, the impaired naturalness and intelligibility of the speech produced by these patients hinder their active participation in societal activities, leading them to social isolation and depression. Speech related assistive aids like speech enhancement and recognition systems can thus greatly benefit the dysarthric speakers, easing their communication process.
>>top<<