Artificial Intelligence from Google to Create Video from Photo [Video]

DeepMind, the artificial intelligence team that was founded in 2010 and joined Google in 2014, introduced a very exciting technology today. Unlike AI tools like DALL-E 2 and MidJourney, the team has developed an AI that can transform an existing thing into something even larger this time around.

DeepMind’s new artificial intelligence, called ‘Transframer’, can create 30-second videos that will allow you to live in the picture based on just a single picture. The resulting video can be saved as a GIF. Let’s take a closer look at what the new artificial intelligence can do.

The area and its surroundings in the photograph are imagined in three dimensions:

Transframer analyzes a photo you upload and estimates the objects in that photo and their surroundings. After this process, artificial intelligence also considers how objects can be seen from different angles. The resulting image is like a video actually shot there.

Here’s how the Transframer creates what an object might look like from different angles:

So what can be done with this technology?

Just like DALL-E 2 and MidJourney, Transframer opens the door to endless options. This artificial intelligence can be used for the further development of art, scientific analysis and artificial intelligence, as well as perhaps transforming the imaginary spaces created with DALL-E 2 and MidJourney into a 360-degree space. These areas can also take place in our lives as areas where we can be in virtual reality in the future. What can be done with artificial intelligence will be limited only by our imagination.

Although artificial intelligence is not yet available to everyone, the DeepMind team shared the algorithms that make this technology possible in their research. Software enthusiasts can access these algorithms and their explanations by clicking this link.