Sber taught its Kandinsky neural network to generate animated videos based on text descriptions. The press service of the technology company spoke about the new capabilities of the technology.
Kandinsky 2.2 users will be able to receive a four-second animated video based on a text description. Video length: 640×640 pixels, frame rate: 24 per second. One second of video is generated on average 20 seconds.
Technically, the process works as follows: the user describes in words what he wants to see in the video, after which the robot offers a selection of 16 options for animating the scene. Composite scene generation is also available. That is, you can configure up to three text descriptions and specify different animation mechanics for each video. In this case, the bot will generate a “mini-movie”.
At an initial stage, the feature will be available to the most active users of Kandinsky 2.2, they will receive corresponding invitations and will be able to test new features. You can apply to get familiar with the new functionality in the Telegram bot.
Previously, Sberbank stated that neural network developers should not be responsible for the images generated by Kandinsky. This was preceded by the deputy’s claim that Kandinsky’s images create a “negative image of Russia.”
Author:
Natalia Gormaleva
Source: RB

I am a professional journalist and content creator with extensive experience writing for news websites. I currently work as an author at Gadget Onus, where I specialize in covering hot news topics. My written pieces have been published on some of the biggest media outlets around the world, including The Guardian and BBC News.