M Chen, K Huang, T Zhao… - … Conference on Machine …, 2023 - proceedings.mlr.press
Diffusion models achieve state-of-the-art performance in various generation tasks. However, their theoretical foundations fall far behind. This paper studies score approximation …
A class of generative models that unifies flow-based and diffusion-based methods is introduced. These models extend the framework proposed in Albergo & Vanden-Eijnden …
We provide the first polynomial-time convergence guarantees for the probabilistic flow ODE implementation (together with a corrector step) of score-based generative modeling. Our …
S Chen, G Daras, A Dimakis - International Conference on …, 2023 - proceedings.mlr.press
We develop a framework for non-asymptotic analysis of deterministic samplers used for diffusion generative modeling. Several recent works have analyzed stochastic samplers …
K Oko, S Akiyama, T Suzuki - International Conference on …, 2023 - proceedings.mlr.press
While efficient distribution learning is no doubt behind the groundbreaking success of diffusion modeling, its theoretical guarantees are quite limited. In this paper, we provide the …
Diffusion models are a powerful method for generating approximate samples from high- dimensional data distributions. Several recent results have provided polynomial bounds on …
Recent works have shown that diffusion models can learn essentially any distribution provided one can perform score estimation. Yet it remains poorly understood under what …
We explore the methodology and theory of reward-directed generation via conditional diffusion models. Directed generation aims to generate samples with desired properties as …
Image denoising—removal of additive white Gaussian noise from an image—is one of the oldest and most studied problems in image processing. Extensive work over several …