Google researchers have developed a new artificial intelligence system called VLOGGER that can generate realistic videos of people speaking, gesturing, and moving based on a single photograph. The technology relies on advanced machine learning models to synthesize surprisingly realistic images, opening up a range of potential applications while also raising concerns about “deepfake” videos.

The VLOGGER AI model from Google can produce highly realistic results using a person’s photograph and a voice recording as input. The model can articulate words to match the person’s voice and generate corresponding hand and facial movements. While the videos are not perfect, they represent a significant milestone in animating static images.

Led by Enric Corona from Google Research, the researchers utilized diffusion models in VLOGGER, a type of machine learning model that has recently been used to generate realistic images from text prompts. By applying these models to the video domain and training them on a large dataset, the team successfully created an AI system that can convincingly bring photos to life.

Google just unveiled VLOGGER, a tool that generates lifelike talking videos with a single photo

Now, anybody can become a YouTuber.

Here’s what you need to know: pic.twitter.com/gpJVBNoAp3

— EyeingAI (@EyeingAI) March 18, 2024

According to the authors of the research, VLOGGER does not require individual training for each person, relies on face detection and cropping, and generates the entire image at once, unlike previous methods. Trained on over 800,000 different identities and 2,200 hours of video, VLOGGER, named MENTOR, could potentially be used in dubbing and video editing. This technology can also be used to create photorealistic avatars for virtual reality and games and may facilitate the creation of more engaging AI-powered virtual assistants and chatbots.

However, along with its potential for positive applications, this technology also poses risks of misuse, such as creating deepfakes where someone’s likeness is replaced with another’s in a video. As AI-generated videos become more realistic and easier to create, the challenge of misinformation and digital forgery becomes a more significant issue. For this reason, VLOGGER is currently only available as a research model.

Stay plugged into the gaming pulse, and anticipate the thrill of new adventures and the nostalgia of departing classics. For the freshest updates, keep an eye on Lavegagames.com – where gaming news comes to life.