Boomerang: Local sampling on image manifolds using diffusion models

arXiv:2210.12100v2 Announce Type: replace-cross
Abstract: The inference stage of diffusion models can be seen as running a reverse-time diffusion stochastic differential equation, where samples from a Gaussian latent distribution are transformed into samples from a target distribution that usually reside on a low-dimensional manifold, e.g., an image manifold. The intermediate values between the initial latent space and the image manifold can be interpreted as noisy images, with the amount of noise determined by the forward diffusion process noise schedule. We …

abstract arxiv cs.cv cs.lg differential differential equation diffusion diffusion models distribution equation image inference intermediate low manifold running samples sampling stage stat.ml stochastic type values