Research map / Image generation / Fast diffusion sampling
Efficient Diffusion Models: A Comprehensive Survey From Principles to Practices
Zhiyuan Ma, Yuzhu Zhang, Guoli Jia, Liangliang Zhao, Yichao Ma, Mingjie Ma, Gaofeng Liu, Kaiyan Zhang, Ning Ding, Jianjun Li, Bowen Zhou
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| increase sampling NFEs from 1 to 4 | DDIM · CIFAR-10 | FID: 13.68 -> 4.67 | Improved |
| increase sampling NFEs from 1 to 4 | PNDM · CIFAR-10 | FID: 7.05 -> 3.68 | Improved |
| increase sampling NFEs from 1 to 4 | DPM-Solver · LSUN | FID: 6.10 -> 2.53 | Improved |
| increase sampling NFEs from 1 to 4 | PNDM · LSUN | FID: 8.69 -> 9.89 | Hurt |
| increase sampling NFEs from 1 to 4 | gDDIM · CIFAR-10 | FID: 41.7 -> 2.59 | Improved |
| increase sampling NFEs from 1 to 4 | Progressive Distillation · CIFAR-10 | FID: 9.12 -> 3.00 | Improved |
| increase sampling NFEs from 1 to 2 | Consistency Models · CIFAR-10 | FID: 3.55 -> 2.93 | Improved |
| increase sampling NFEs from 1 to 4 | TRACT · CIFAR-10 | FID: 3.78 -> 2.93 | Improved |
| increase sampling NFEs from 1 to 2 | DDGAN · CIFAR-10 | FID: 14.60 -> 4.08 | Improved |
| distill diffusion into one-step generator (DMD) versus DDIM at 1 NFE | DMD · CIFAR-10 | FID: 13.68 -> 2.62 | Improved |
| use ODE solver DPM-Solver instead of DDIM at 1 NFE | DPM-Solver · CIFAR-10 | FID: 13.68 -> 6.37 | Improved |
| increase NFEs from 100 to 200 in truncated diffusion | ES-DDPM · CIFAR-10 | FID: 5.52 -> 5.02 | Improved |
Closest papers by meaning
- A Simple Early Exiting Framework for Accelerated Sampling in Diffusion Models (ICML 2024 · similarity 0.91)
- Multi-Architecture Multi-Expert Diffusion Models (AAAI 2024 · similarity 0.90)
- Align Your Steps: Optimizing Sampling Schedules in Diffusion Models (ICML 2024 · similarity 0.90)
- Simple and Fast Distillation of Diffusion Models (NeurIPS 2024 · similarity 0.90)
- Diffusion Model is Effectively Its Own Teacher (CVPR 2025 · similarity 0.90)
- Revisiting Diffusion Models: From Generative Pre-training to One-Step Generation (ICML 2025 · similarity 0.90)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
