A latent diffusion model (LDM) is a generative model that performs the diffusion process in a lower-dimensional latent space, rather than directly in pixel space. This approach significantly reduces computational costs while maintaining high-fidelity image generation capabilities.
Latent diffusion models are a type of AI that generates images by working with a compressed version of the image rather than the full pixel data. This makes them much faster and more efficient, allowing for the creation of high-quality images with less computational power. They are particularly useful as generative priors for tasks like image compression.
LDM, Stable Diffusion (a prominent LDM), Text-to-Image LDM
Was this definition helpful?