Research map / Image generation / Fast diffusion sampling
One-shot Conditional Sampling: MMD meets Nearest Neighbors
Anirban Chatterjee, Sayantan Choudhury, Rohan Hore
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 |
|---|---|---|---|
| use CGMMD instead of full diffusion model for conditional sampling | MNIST denoising · MNIST denoising (sigma=0.9) | time per image (s): 5.42 × 10−2 -> 5.6 × 10−4 | Improved |
| use CGMMD instead of full diffusion model for conditional sampling | MNIST denoising · MNIST denoising (sigma=0.9) | PSNR: 13.326 -> 8.922 | Hurt |
| use CGMMD instead of distilled diffusion | MNIST denoising · MNIST denoising (sigma=0.9) | SSIM: 0.508 -> 0.718 | Improved |
| use CGMMD instead of GCDS and WGAN at high noise | Helix conditional sampling · Helix benchmark, sigma 0.2 to 0.6 | qualitative sample fidelity | Improved |
| add l1 regularisation to WGAN | WGAN baseline · Helix benchmark | training stability | Improved |
| increase ambient dimension d from 25 to 500 with fixed intrinsic dimension 5 (ablation) | CGMMD · Synthetic intrinsic-dimension experiment | MSE of conditional mean and SD | No clear effect |
| vary kernel among Gaussian, Laplace and IMQ (ablation) | CGMMD · Synthetic intrinsic-dimension experiment | MSE of conditional mean and SD | No clear effect |
Closest papers by meaning
- A Black-Box Debiasing Framework for Conditional Sampling (NeurIPS 2025 · similarity 0.87)
- Posterior Sampling Based on Gradient Flows of the MMD with Negative Distance Kernel (ICLR 2024 · similarity 0.87)
- CADS: Unleashing the Diversity of Diffusion Models through Condition-Annealed Sampling (ICLR 2024 · similarity 0.87)
- A Simple Approach to Unifying Diffusion-based Conditional Generation (ICLR 2025 · similarity 0.86)
- Theory on Score-Mismatched Diffusion Models and Zero-Shot Conditional Samplers (ICLR 2025 · similarity 0.86)
- NoiseCtrl: A Sampling-Algorithm-Agnostic Conditional Generation Method for Diffusion Models (CVPR 2025 · similarity 0.86)
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.
