Explore the transformative capabilities of diffusion models in image generation during this 30-minute conference talk from DSC EUROPE 24 in Belgrade. Learn about their applications in popular tools like Midjourney, DALL-E 3, and Stable Diffusion, and discover how these models excel at generating high-quality images with flexibility for inpainting, remixing, and image animation. Gain valuable insights into the typical requirements for training large diffusion models and understand common pitfalls encountered during development. This practical knowledge, presented by Eugene Lyapustin on November 20th, will help you fully harness the potential of diffusion models in both creative and technical domains.