AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Formulating Discrete Probability Flow Through Optimal Transport

Pengze Zhang, Hubery Yin, Chen Li, Xiaohua Xie

NeurIPS 2023 · Image generation › Fast diffusion sampling · approach: Diffusion theory and learned samplers

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 changedWhereMeasuredEffect
replace SDDM reverse generator with discrete probability flow generator (DPF)SDDM · Synthetic 2spirals, dim 32 state 2CSD: 14.3053 -> 2.1719Improved
use DPF generator instead of SDDMSDDM · Synthetic 8gaussians, dim 32 state 2CSD: 14.1882 -> 1.7945Improved
use DPF generator instead of SDDMSDDM · Synthetic checkerboard, dim 32 state 2CSD: 14.7433 -> 2.0693Improved
use DPF generator instead of SDDM, state size 5SDDM · Synthetic 2spirals, dim 16 state 5CSD: 14.4645 -> 1.9711Improved
use DPF generator instead of SDDM, state size 5SDDM · Synthetic checkerboard, dim 16 state 5CSD: 14.6963 -> 1.4172Improved
use DPF generator instead of SDDM, state size 10SDDM · Synthetic 2spirals, dim 12 state 10CSD: 12.8463 -> 1.8123Improved
use DPF generator instead of SDDM, state size 10SDDM · Synthetic checkerboard, dim 12 state 10CSD: 13.0158 -> 1.1348Improved
use DPF generator instead of SDDMSDDM · Synthetic 2spirals, dim 32 state 2MMD (lower is better): 2.18e-06 -> 1.89e-05Hurt
use DPF generator instead of SDDMSDDM · Synthetic circles, dim 32 state 2MMD (lower is better): 6.22e-06 -> 3.27e-05Hurt
use DPF generator instead of SDDM, state size 5SDDM · Synthetic 8gaussians, dim 16 state 5MMD (lower is better): 1.01e-4 -> 5.87e-4Hurt
use DPF generator instead of SDDM, state size 10SDDM · Synthetic 2spirals, dim 12 state 10MMD (lower is better): 5.52e-4 -> 7.19e-4Hurt
use DPF generator instead of SDDMSDDM · Synthetic data, all three state sizessample quality (visual)No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy