Speech Replacement
Project Date
May, 2020
Techniques
Audio Analysis, Deep Learning, Feature Engineering
Visit
GitHub Repository
By taking the data extracted in an audio file, the idea of this project was to replace the voice recording of a person with the voice from a speaker used during the model training process. The goal was to achieve results similar to Auto-Tune, in which the user’s voice would be displaced or modified to resemble the voice of another speaker.
However, the approach used in this project did not result in any usable audio output.