Diffusion Rejection Sampling

TL;DR

Diffusion Rejection Sampling (DiffRS) is a novel method that improves diffusion model sampling accuracy by using a rejection scheme to refine samples at each step, achieving state-of-the-art results.

Contribution

The paper introduces DiffRS, a rejection sampling approach that aligns transition kernels with true distributions, enhancing sampling quality and efficiency in diffusion models.

Findings

Achieves tighter sampling error bounds.

Demonstrates state-of-the-art performance on benchmark datasets.

Effective for fast diffusion and large-scale text-to-image models.

Abstract

Recent advances in powerful pre-trained diffusion models encourage the development of methods to improve the sampling performance under well-trained diffusion models. This paper introduces Diffusion Rejection Sampling (DiffRS), which uses a rejection sampling scheme that aligns the sampling transition kernels with the true ones at each timestep. The proposed method can be viewed as a mechanism that evaluates the quality of samples at each intermediate timestep and refines them with varying effort depending on the sample. Theoretical analysis shows that DiffRS can achieve a tighter bound on sampling error compared to pre-trained models. Empirical results demonstrate the state-of-the-art performance of DiffRS on the benchmark datasets and the effectiveness of DiffRS for fast diffusion samplers and large-scale text-to-image diffusion models. Our code is available at https://github.com/aailabkaist/DiffRS.