Kandinsky 3.0 Technical Report
This work addresses the need for more efficient and higher-quality image generation tools for users in creative and AI applications, though it appears incremental as part of an ongoing series.
The authors tackled the problem of improving text-to-image generation quality and realism with Kandinsky 3.0, a latent diffusion model, achieving better text understanding and specific domain performance through human preference comparisons, and introduced a distilled version, Kandinsky 3.1, that is 20 times faster without quality loss.
We present Kandinsky 3.0, a large-scale text-to-image generation model based on latent diffusion, continuing the series of text-to-image Kandinsky models and reflecting our progress to achieve higher quality and realism of image generation. In this report we describe the architecture of the model, the data collection procedure, the training technique, and the production system for user interaction. We focus on the key components that, as we have identified as a result of a large number of experiments, had the most significant impact on improving the quality of our model compared to the others. We also describe extensions and applications of our model, including super resolution, inpainting, image editing, image-to-video generation, and a distilled version of Kandinsky 3.0 - Kandinsky 3.1, which does inference in 4 steps of the reverse process and 20 times faster without visual quality decrease. By side-by-side human preferences comparison, Kandinsky becomes better in text understanding and works better on specific domains. The code is available at https://github.com/ai-forever/Kandinsky-3