High-Resolution Image Synthesis with Latent Diffusion Models
Talk, Universidad de los Andes, Biomedical Computer Vision Group, Bogotá, Colombia
In this seminar, I presented the paper “High-Resolution Image Synthesis with Latent Diffusion Models” by Rombach et al., from Ludwig Maximilian University of Munich, IWR at Heidelberg University, and Runway ML. This influential work, accepted at CVPR 2022, introduces a novel approach to high-quality image generation by performing diffusion in a compressed latent space, enabling both efficiency and state-of-the-art synthesis results.