Kandinsky 3: Text-to-Image Synthesis for Multifunctional Generative Framework
This work provides a multifunctional generative framework for users in creative and AI applications, though it is incremental as it builds upon existing latent diffusion models.
The authors tackled the problem of developing a versatile text-to-image diffusion model by introducing Kandinsky 3, which achieves high-quality and photorealistic results while enabling adaptation to multiple generation tasks such as inpainting, image fusion, and video generation, with a distilled version offering 3x faster inference without quality loss.
Text-to-image (T2I) diffusion models are popular for introducing image manipulation methods, such as editing, image fusion, inpainting, etc. At the same time, image-to-video (I2V) and text-to-video (T2V) models are also built on top of T2I models. We present Kandinsky 3, a novel T2I model based on latent diffusion, achieving a high level of quality and photorealism. The key feature of the new architecture is the simplicity and efficiency of its adaptation for many types of generation tasks. We extend the base T2I model for various applications and create a multifunctional generation system that includes text-guided inpainting/outpainting, image fusion, text-image fusion, image variations generation, I2V and T2V generation. We also present a distilled version of the T2I model, evaluating inference in 4 steps of the reverse process without reducing image quality and 3 times faster than the base model. We deployed a user-friendly demo system in which all the features can be tested in the public domain. Additionally, we released the source code and checkpoints for the Kandinsky 3 and extended models. Human evaluations show that Kandinsky 3 demonstrates one of the highest quality scores among open source generation systems.