Student: Sneha Raman
Supervisor: Eva Navas Cordón e Inma Hernáez Rioja
The aim of this project is to use a voice conversion methodology that would enrich alaryngeal speech by improving its intelligibility and maintaining speaker identity. We also propose to measure listening effort and cognitive load using pupillometry, heart rate measurement and EEG, before enrichment and at different stages of enrichment of the speech. The major accomplishments expected are:
• Implementing eigenvoice voice conversion for pathological voices.
• Implementing voice conversion based on deep neural networks.
• Measuring cognitive effort associated with pathological speech as a metric to track the improvement in intelligibility.