Learns to reverse a gradual noising process, generating by denoising repeatedly from pure noise. Trades sampling speed for stable training and far better mode coverage than adversarial methods.
supersedescorrects · extends
classifiesspecializes · part-of
substitutes forapproximates · alternative-to
depends onrequires · validates
Colour is the family; a dashed line is the second member of it.
Drag to pan · scroll to zoom · click a node to open it
This node
alternative-tois a competing approach to Generative Adversarial Networka stable objective instead of a minimax game that collapses
extendsadds capability to Learned Denoisingdenoises repeatedly, from pure noise, to synthesise
Referenced by
correctsClassifier-Free Guidance fixes a defect in thisconditioning alone is too weak to follow a prompt
correctsLatent Diffusion fixes a defect in thisdenoising at pixel resolution is prohibitively expensive