Machine Learning-Based Mask Removal
Digital Domain a behind-the-scenes look at their new toolkit.
The team of Digital Domain presented their study of mask removal launched a couple of months back. They wanted to explore the idea of using Charlatan (the team's proprietary neural renderer) to remove face masks on close-up subjects. They were happy with the results and decided to continue the research and create a facial performance underneath the mask using only an audio track.
"What you see here is still very much a work in progress but it is exciting to see where this technology is going," wrote the team. "There is very little manual artist work in this piece and every time you see maskless Doug everything on his face from his eyes to his mouth is synthetic!"
And here's a look at a different project from the team, an autonomous digital human called Douglas:
Learn more about the team here. Don't forget to join our new Telegram channel, our Discord, follow us on Instagram and Twitter, where we are sharing breakdowns, the latest news, awesome artworks, and more.
