Synthesizing Obama: Learning Lip Sync from Audio

Created on 2022-05-08T02:55:43-05:00

Return to the Index

This card pertains to a resource available on the internet.

This card can also be read via Gemini.

Uses LSTMs to predict mouth shapes from audio.

They don't produce a mouth so much as a "mouth PCA" feature set, which is a set of variables created from doing principal component analysis on the actual mouth feature set.