AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Innovating Real Fisheye Image Correction with Dual Diffusion Architecture

Shangrong Yang, Chunyu Lin, Kang Liao, Yao Zhao

ICCV 2023 · Image generation › Fast diffusion sampling · approach: Denoising diffusion probabilistic models · 17 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 one-pass network (OPN) to generate a new condition for the conditional diffusion module (ablation)CDM · Synthetic fisheye, inferencePSNR / FID: PSNR 18.6 -> 22.1, FID 127.9 -> 55.1Improved
add unconditional diffusion module trained on unlabeled real fisheye images (ablation)CDM+OPN · Synthetic fisheye, inferencePSNR / FID: PSNR 22.1 -> 24.6, FID 55.1 -> 24.9Improved
add unconditional diffusion module (UDM) on real images (ablation)CDM+OPN · Synthetic fisheye, one-passPSNR / FID: PSNR 25.9 -> 26.0, FID 56.4 -> 57.8No clear effect
CDM+OPN without UDM on real fisheye images (ablation)CDM+OPN · Woodscape real fisheyevisual artifactsHurt
use iterative diffusion inference instead of one-passDDA · Synthetic fisheyePSNR / SSIM vs FID / LPIPS: PSNR 26.0 -> 24.6, FID 57.8 -> 24.9Mixed
use DDA instead of PCNfisheye correction · Synthetic fisheyeFID: 65.8 -> 24.9Improved

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