Join Coding Elite (get early access to new videos and projects of mine, and request for my models to try a new text/song): http://codingelite.com/join
Another attempt at generating a Donald Trump singing voice. 100% of this audio was generated by my improved Trump text-to-speech model (trained from an implementation of the paper "Style Tokens: Unsupervised Style Modeling, Control and Transfer in End-to-End Speech Synthesis" (https://arxiv.org/pdf/1803.09017.pdf)), with no post-processing or auto tune effects added. Given a line to say and a reference audio, the model will try to say that line in Trump's voice in the style of the reference audio. Mouth movements were generated with my own method for transferring facial expressions (see https://www.youtube.com/watch?v=1bB5d... for a better example - obviously here the lip sync is very far off and could be vastly better).