Emotional-Speech-Synthesis

This work was done for the course DT2119 Speech and Speaker Recognition (2022) at Royal Institute of Technology (KTH), Stockholm. Read the report.pdf for all information about the project.

The project aimed to condition a TTS system (Tacotron2 + WaveGlow) on a emotion or a speaker identity using the IEMOCAP and the VCTK datasets. It builds upon the following repositories:
https://github.com/NVIDIA/tacotron2
https://github.com/NVIDIA/waveglow
https://github.com/resemble-ai/Resemblyzer