AI Paper Atlas

Research map / Image generation / Image restoration

Real-World Blur Dataset for Learning and Benchmarking Deblurring Algorithms

Jaesung Rim, Haeyun Lee, Jucheol Won, Sunghyun Cho

ECCV 2020 · Image generation › Image restoration · approach: Image deblurring · 426 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 blur-kernel-based alignment to reference homography (ablation)dataset alignment · RealBlur alignmentdisplacement error (pixels): 4.9454 -> 0.8701Improved
add phase correlation to homography plus blur kernel alignment (ablation)dataset alignment · RealBlur alignmentdisplacement error (pixels): 0.8701 -> 0.8058Improved
align data with blur-kernel centroids instead of phase correlation (ablation)SRN-DeblurNet · Kohler et al. testPSNR (dB): 26.48 -> 28.07Improved
align data with centroids instead of non-zero centers (ablation)SRN-DeblurNet · Kohler et al. testPSNR (dB): 27.80 -> 28.07Improved
train on RealBlur-J instead of GoPro (ablation)SRN-DeblurNet · RealBlur-J testPSNR (dB): 28.56 -> 31.02Improved
train on RealBlur-R instead of GoPro (ablation)SRN-DeblurNet · RealBlur-R testPSNR (dB): 35.66 -> 36.47Improved
train on BSD-B instead of GoPro (ablation)SRN-DeblurNet · RealBlur-R testPSNR (dB): 35.66 -> 34.96Hurt
combine all training sets with pre-trained weights versus RealBlur-R only (ablation)SRN-DeblurNet · RealBlur-R testPSNR (dB): 36.47 -> 38.65Improved
combine all training sets with pre-trained weights versus RealBlur-J only (ablation)SRN-DeblurNet · RealBlur-J testPSNR (dB): 31.02 -> 31.38Improved
train SRN-DeblurNet on RealBlur data instead of released pretrained modelSRN-DeblurNet · RealBlur-J testPSNR (dB): 28.56 -> 31.38Improved
train SRN-DeblurNet on RealBlur data instead of released pretrained modelSRN-DeblurNet · RealBlur-R testPSNR (dB): 35.66 -> 38.65Improved
train DeblurGAN-v2 on RealBlur data instead of released pretrained modelDeblurGAN-v2 · RealBlur-J testPSNR (dB): 28.70 -> 29.69Improved

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