Google's New VLOGGER AI Is Next-Level Deepfake Video Generator
VLOGGER is a novel framework that can generate human-like vlogs with an audio and a photo as input.
AI is moving from simple image deepfakes to videos.
No, I am not talking about swapping faces like the typical deepfake videos we’ve seen before—it’s something far more intriguing and potentially unsettling.
Today, Google released a research paper detailing a novel framework called VLOGGER that lets you generate a video of a human vlogger using only an audio clip and a single image as input.
If you thought deepfakes were scary, this takes the use of AI technology to a whole new level.
What is VLOGGER?
VLOGGER uses a multimodal diffusion technique to synthesize humans from audio. It can generate photorealistic videos of a person talking with realistic head movements, facial expressions, gazes, and even hand gestures.
Here’s an example:
Keep reading with a 7-day free trial
Subscribe to Generative AI Publication to keep reading this post and get 7 days of free access to the full post archives.