Subspace Diffusion Generative Models
arXiv:2205.01490 · doi:10.1007/978-3-031-20050-2_17
Abstract
Score-based models generate samples by mapping noise to data (and vice versa) via a high-dimensional diffusion process. We question whether it is necessary to run this entire process at high dimensionality and incur all the inconveniences thereof. Instead, we restrict the diffusion via projections onto subspaces as the data distribution evolves toward noise. When applied to state-of-the-art models, our framework simultaneously improves sample quality -- reaching an FID of 2.17 on unconditional CIFAR-10 -- and reduces the computational cost of inference for the same number of denoising steps. Our framework is fully compatible with continuous-time diffusion and retains its flexible capabilities, including exact log-likelihoods and controllable generation. Code is available at https://github.com/bjing2016/subspace-diffusion.
ECCV 2022
Cited by in corpus (6)
- Diffusion Models in Vision: A Survey
- StoRM: A Diffusion-based Stochastic Regeneration Model for Speech Enhancement and Dereverberation
- Diffusion Models, Image Super-Resolution And Everything: A Survey
- Dehazing Ultrasound using Diffusion Models
- Fast Graph Generation via Spectral Diffusion
- Sequential Posterior Sampling with Diffusion Models