Pith. sign in

REVIEW 16 cited by

Fast Sampling of Diffusion Models with Exponential Integrator

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2204.13902 v4 pith:YGR6YR4B submitted 2022-04-29 cs.LG

Qinsheng Zhang , Yongxin Chen This is my paper · ORCID
classification cs.LG
keywords diffusionsamplingdiscretizationexponentialintegratorlearnedmethodprocess
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
abstract

The past few years have witnessed the great success of Diffusion models~(DMs) in generating high-fidelity samples in generative modeling tasks. A major limitation of the DM is its notoriously slow sampling procedure which normally requires hundreds to thousands of time discretization steps of the learned diffusion process to reach the desired accuracy. Our goal is to develop a fast sampling method for DMs with a much less number of steps while retaining high sample quality. To this end, we systematically analyze the sampling procedure in DMs and identify key factors that affect the sample quality, among which the method of discretization is most crucial. By carefully examining the learned diffusion process, we propose Diffusion Exponential Integrator Sampler~(DEIS). It is based on the Exponential Integrator designed for discretizing ordinary differential equations (ODEs) and leverages a semilinear structure of the learned diffusion process to reduce the discretization error. The proposed method can be applied to any DMs and can generate high-fidelity samples in as few as 10 steps. In our experiments, it takes about 3 minutes on one A6000 GPU to generate $50k$ images from CIFAR10. Moreover, by directly using pre-trained DMs, we achieve the state-of-art sampling performance when the number of score function evaluation~(NFE) is limited, e.g., 4.17 FID with 10 NFEs, 3.37 FID, and 9.74 IS with only 15 NFEs on CIFAR10. Code is available at https://github.com/qsh-zh/deis

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 16 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Asymptotic Preservation and Uniform Accuracy of Diffusion and Flow-Matching Samplers

    cs.LG 2026-07 conditional novelty 7.5 of 10

    DDIM (σ-clock Euler) is the unique layer-exact fixed-step sampler; deterministic residual budgets stay O(1) with no log(1/σ_min), while stochastic path-KL scales as Λ²/N from the Itô term alone.

  2. Blade: A Derivative-free Bayesian Inversion Method using Diffusion Priors

    cs.LG 2025-10 conditional novelty 7.0 of 10

    Blade produces well-calibrated derivative-free Bayesian posterior samples by combining split Gibbs sampling, ensemble statistical linearization, and diffusion priors.

  3. Amortized Moment Matching for Visual Generation

    cs.LG 2026-07 accept novelty 6.0 of 10

    Amortized Fréchet Distance uses neural nets to match conditional means and covariances, yielding stronger one-step visual generators than explicit FD-loss or multi-step teachers.

  4. FARI: Robust One-Step Inversion for Watermarking in Diffusion Models

    cs.CR 2026-07 accept novelty 6.0 of 10

    One-step adversarially LoRA-tuned inversion exploits low-curvature reverse trajectories to beat 50-step DDIM on watermark robustness after ~20 minutes of fine-tuning.

  5. Function-Space Diffusion for Motion Planning

    cs.RO 2026-07 accept novelty 6.0 of 10

    FSD-MP performs diffusion over continuous trajectory functions with Matérn noise and a DST-FNO reverse model, enabling zero-shot multi-resolution robot motion planning.

  6. DanceOPD: On-Policy Generative Field Distillation

    cs.CV 2026-06 unverdicted novelty 6.0 of 10

    Hard-routed, single low-noise on-policy velocity matching composes conflicting image-generation capabilities into one flow student better than joint training, merging, or dense OPD baselines.

  7. RSTR: Reducing SpatioTemporal Redundancy in Diffusion Transformers

    cs.CV 2025-12 conditional novelty 6.0 of 10

    OUSAC uses evolutionary search to skip most classifier-free-guidance passes and adaptive low-rank caching to recover quality, cutting diffusion transformer compute by about 50–60% with equal or better FID.

  8. Fast Convergence for High-Order ODE Solvers in Diffusion Probabilistic Models

    cs.LG 2025-06 conditional novelty 6.0 of 10

    A TV convergence bound O(d^{7/4} ε^{1/2} + d(dH)^p) is proved for p-th order (exponential) Runge-Kutta samplers of probability-flow ODEs under C² smoothness of the learned score.

  9. Diffusion models under low-noise regime

    cs.CV 2025-06 conditional novelty 6.0 of 10

    Diffusion models trained on disjoint data converge at high noise but diverge near the data manifold, and they fail to denoise very small perturbations accurately.

  10. MaRS: A Fast Sampler for Mean Reverting Diffusion based on ODE and SDE Solvers

    cs.CV 2025-02 conditional novelty 6.0 of 10

    MR Sampler derives semi-analytical solutions to the reverse-time SDE and PF-ODE of Mean Reverting Diffusion, enabling 5-10 step high-quality sampling instead of hundreds.

  11. CHORDS: Diffusion Sampling Accelerator with Multi-core Hierarchical ODE Solvers

    cs.LG 2025-07 conditional novelty 5.0 of 10

    CHORDS accelerates diffusion sampling by running hierarchical ODE solvers on multiple cores, with slower solvers rectifying faster ones, achieving up to 2.9x speedup without retraining.

  12. Straighten Viscous Rectified Flow via Noise Optimization

    cs.CV 2025-07 reject novelty 5.0 of 10

    VRFNO claims state-of-the-art one- and few-step image generation by straightening rectified flow trajectories, but its sampler relies on real images from the dataset.

  13. Unleashing High-Quality Image Generation in Diffusion Sampling Using Second-Order Levenberg-Marquardt-Langevin

    cs.CV 2025-05 reject novelty 5.0 of 10

    A training-free 'Levenberg-Marquardt-Langevin' diffusion sampler is claimed to improve image FID, but its update rule collapses to that of the baseline DPM-Solver for the parameter values used in the paper.

  14. FSampler: Training Free Acceleration of Diffusion Sampling via Epsilon Extrapolation

    cs.LG 2025-11 conditional novelty 4.0 of 10

    FSampler accelerates diffusion sampling by substituting finite-difference extrapolations of epsilon for model calls on selected steps, reducing NFE by 15-25% at SSIM 0.95-0.99.

  15. Censored Sampling for Topology Design: Guiding Diffusion with Human Preferences

    cs.LG 2025-08 unverdicted novelty 4.0 of 10

    Guiding a pretrained topology-diffusion generator with human-preference reward classifiers is claimed to suppress floating-material and boundary-violation failure modes without retraining the generator.

  16. DualFast: Dual-Speedup Framework for Fast Sampling of Diffusion Models

    cs.CV 2025-06 conditional novelty 4.0 of 10

    A training-free correction that blends each step's noise estimate with the initial noise estimate improves few-step diffusion sampling across DDIM, DPM-Solver, and DPM-Solver++.

Pith tools