Generate expressive character performances with video inputs
Act-One is a new state-of-the-art tool for generating expressive character performances using video and voice performances as inputs. It represents a significant step forward in using generative models for expressive live action and animated content.