Experience

Speech Enhancement/Separation

Take me to pookie

An important question in developing DNN-based speech enhancement systems for CI users is, what optimization criterion should be used to train the networks? Mean square error (MSE) or log-MSE measures are normally used, but research in CI domain has shown that these measures do not correlate well with the speech intelligibility of CI users. Conditional generative adversarial network (CGAN) provides a solution by replacing the optimization loss functions with a discriminator network. In this project, I am exploring different methods of leveraging CGAN-based architectures to improve the speech intelligibility of CI users.

Emotion Recognition

Mood Recognition

Statistical Parametric Speech Synthesis

Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis —————————————

this is pookie

Statistical Parametric Speech Synthesis

Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis ————————————— Statistical Parametric Speech Synthesis —————————————

Text-To-Speech