Face2Face: Real-Time Face Capture and Reenactment of RGB Videos
niessnerlab.org
niessnerlab.org
On an earlier Tacotron article: https://news.ycombinator.com/item?id=15963888
https://google.github.io/tacotron/publications/uncovering_la...
Seems to do with manipulating the speed styles in how we speak.
Listen to the "neutral" blue lagoon, vs token 6 (tired sound), vs mix w/ token 7 (prominence)" (very confident).
GANs should be able to do a lot better now.