Accelerating Guided Diffusion Sampling with Splitting Numerical Methods
This work addresses a key bottleneck in diffusion models for practitioners needing faster conditional generation in tasks like text-to-image and inpainting, though it is incremental as it builds on existing acceleration techniques.
The paper tackled the problem of slow guided diffusion sampling by identifying that classical high-order numerical methods fail for conditional functions, and proposed an operator splitting method that reduces sampling time by 32-58% while maintaining image quality on ImageNet256.
Guided diffusion is a technique for conditioning the output of a diffusion model at sampling time without retraining the network for each specific task. One drawback of diffusion models, however, is their slow sampling process. Recent techniques can accelerate unguided sampling by applying high-order numerical methods to the sampling process when viewed as differential equations. On the contrary, we discover that the same techniques do not work for guided sampling, and little has been explored about its acceleration. This paper explores the culprit of this problem and provides a solution based on operator splitting methods, motivated by our key finding that classical high-order numerical methods are unsuitable for the conditional function. Our proposed method can re-utilize the high-order methods for guided sampling and can generate images with the same quality as a 250-step DDIM baseline using 32-58% less sampling time on ImageNet256. We also demonstrate usage on a wide variety of conditional generation tasks, such as text-to-image generation, colorization, inpainting, and super-resolution.