Alibaba's EMO AI generates lifelike actors and singers
Summary: Alibaba's Institute for Intelligent Computing released EMO, an AI video generator that turns still images into lifelike actors and singers. EMO's demo includes Sora singing a Dua Lipa song and Audrey Hepburn speaking Lili Reinhart's audio. EMO outperforms NVIDIA's Audio2Face in emoting and facial realism. EMO's model uses a large dataset and a diffusion-based approach without 3D models. It supports multiple languages and accurately emotes during pauses. The system's capabilities are impressive, but its impact on the acting industry is concerning.
The article metrics are deprecated.
I'm replacing the original 8-factor scoring system with a new and improved one. It doesn't use the original factors and gives much better significance scores.