Visar

Assistant Professor

Electrical, Computer, and Energy Engineering
Speech and Hearing Science
Arizona State University
COOR 3472
(480) 727 - 6455
visar ((at)) asu ((dot)) edu

Reducing the complexity of deep learning architectures (PhD or Postdoc)

We are looking for PhD students or postdoctoral researchers interested in developing new algorithms for reducing the complexity of deep learning architectures, while preserving their ability to model complex data. The ideal applicant will have a background in fundamental statistical signal processing and information theory concepts and an interest in deep learning. If you are interested, please send a CV with references to Prof. Berisha.

Computational modeling of prosody and articulation in populations with neurogenic disorders (PhD or Postdoc)

Neurological disorders negatively impact the speech production mechanism and, as a result, the produced speech has atypical rhythm, stress, and articulation. In this project our aim is to model the relationship between disturbances in dysarthric speech and perceived intelligibility at multiple scales (segmental to suprasegmental). This requires 1) developing novel acoustic representations of atypical prosody and reduced articulatory precision, 2) characterizing the relationship between these acoustic measures and intelligibility scores. We are seeking motivated Postdoctoral Researchers or PhD students to fill these positions. The ideal applicant will have a background in speech signal processing, applied machine learning, and will be excited about the possibility of working with clinical populations. If you are interested please send a CV with references to Prof. Berisha.