AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

BokehDiff: Neural Lens Blur with One-Step Diffusion

Chengxuan Zhu, Qingnan Fan, Qi Zhang, Jinwei Chen, Huaqi Zhang, Chao Xu, Boxin Shi

ICCV 2025 · Image generation › Fast diffusion sampling · approach: Diffusion posterior sampling · 4 citations (OpenAlex)

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
add adversarial loss (ablation)BokehDiff · EBB Val294 (exposure-aligned)LPIPS: 0.3768 -> 0.3737Improved
add VGG perceptual loss (ablation)BokehDiff · EBB Val294 (exposure-aligned)LPIPS: 0.4218 -> 0.3737Improved
add multi-scale edge loss (ablation)BokehDiff · EBB Val294 (exposure-aligned)LPIPS: 0.3785 -> 0.3737Improved
train the VAE encoder instead of fixing it (ablation)BokehDiff · EBB Val294 (exposure-aligned)PSNR: 24.266 -> 24.652Improved
add circle-of-confusion constraint in PISA (ablation)BokehDiff · EBB Val294 (exposure-aligned)PSNR: 22.217 -> 24.652Improved
add energy-conserved normalization (SoftmaxQ) (ablation)BokehDiff · EBB Val294 (exposure-aligned)LPIPS: 0.3800 -> 0.3737Improved
add self-occlusion in PISA (ablation)BokehDiff · EBB Val294 (exposure-aligned)LPIPS: 0.3808 -> 0.3737Improved
use timestep T=499 instead of 249 (ablation)BokehDiff · EBB Val294 (exposure-aligned)PSNR: 24.646 -> 24.652No clear effect
use timestep T=499 instead of 749 (ablation)BokehDiff · EBB Val294 (exposure-aligned)PSNR: 24.481 -> 24.652No clear effect
use BokehDiff instead of BokehMeBokehMe · BLB Level 5 (synthetic)PSNR: 38.802 -> 36.798Hurt
use BokehDiff instead of BokehMeBokehMe · BLB Level 5 (synthetic)LPIPS: 0.1404 -> 0.0888Improved
use BokehDiff instead of BokehMeBokehMe · EBB Val294 (exposure-aligned)PSNR: 24.014 -> 24.652Improved

Later papers that compared with this method

2 later accepted papers report BokehDiff as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.

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