New Articles
Windows 11 users have discovered a funny bug that benefits older computers....
It's easy to turn off the transmission — we tell you how to do it....
Such photos have been taken by models and social media users for a long time,...
A famous musician? A schoolteacher? Mom? Tell us about the people you looked up...
Thanks to the instructions of Artyom Kozoriz, you can cope no worse than a...
5 interesting exercises that will help you develop flexibility....
From "Starship Troopers" and "The Matrix" to...
The return of Garfield and Mufasa, the new Transformers and the Lord of the...
About Everything Wiki » Web Services » Sber added an analog of generative fill from Photoshop to the Kandinsky neural network

Sber added an analog of generative fill from Photoshop to the Kandinsky neural network

22 Jan 2024, 00:05, parser
0 comments    0 Show

Sber has introduced a major update to the Kandinsky generative neural network. With the upgrade to Kandinsky 3.0, the generation of illustrations has significantly improved, and the new Kandinsky Video is the first Russian neural network capable of creating videos.

Let's start with Kandinsky 3.0. The main innovation compared to version 2.2 was improved query recognition: generation now corresponds more accurately to the prompt without compromising quality. In addition, the quality of understanding of queries on the topic of the national cultural code — the heroes of Soviet and Russian films and cartoons - has increased significantly. The difference was clearly shown in queries with Cheburashka and Kuzya the housekeeper:

On the left, the results of generation with Kandinsky 3.0, on the right — with Kandinsky 2.2. Illustrations: "Sber"

They also showed a comparison with older versions of Kandinsky and other popular models — Midjourney (signed as MJv.52), Stable Diffusion XL (SDXL) and DALL-E 3. Here are the results of the generation for the query "beautiful girl":

Illustration: "Sber"

This is a "man with a beard":

Illustration: "Sber"

And this is "Barbie and Ken are shopping":

Illustration: "Sber"

Another innovation is the Inpainting and Outpainting modes — the ability to "fit" a new object into an existing image or finish it (it resembles generative fill in Photoshop). Here is an example of finishing:

Illustration: "Sber"

And these are examples of adding an object.

Illustration: "Sber"

You can try out the neural network on the Fusion Brain platform, in the official Telegram bot or VKontakte.

Kandinsky Video allows you to create small animations based on a text query — up to 8 seconds with a frequency of about 30 frames per second and a resolution of up to 512 pixels on the larger side. The height and width can be set by the user.

Illustration: "Sber"

Video creation is available in beta mode on Fusion Brain, and Telegram bot so far offers only to sign up for the waiting list.

reload, if the code cannot be seen