AI Paper Atlas

Research map / Image generation / Image restoration

Unsupervised Domain-Specific Deblurring via Disentangled Representations

Boyu Lu, Jun-Cheng Chen, Rama Chellappa

CVPR 2019 · Image generation › Image restoration · approach: Image deblurring · 173 citations (OpenAlex)

Check 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 the blurring branch (ablation)Unsupervised domain-specific deblurring model · CelebA testPSNR: 18.83 -> 19.84Improved
add the blurring branch (ablation)Unsupervised domain-specific deblurring model · CelebA testd_VGG: 82.9 -> 65.5Improved
add content and blur disentanglement without KL regularization (ablation)Unsupervised domain-specific deblurring model · CelebA testPSNR: 19.84 -> 19.58Hurt
add the KL divergence loss to the blur encoder (ablation)Unsupervised domain-specific deblurring model · CelebA testPSNR: 19.58 -> 20.29Improved
add the KL divergence loss to the blur encoder (ablation)Unsupervised domain-specific deblurring model · CelebA testd_VGG: 69.8 -> 60.6Improved
add the perceptual loss (ablation)Unsupervised domain-specific deblurring model · CelebA testd_VGG: 60.6 -> 57.6Improved
use the perceptual lossUnsupervised domain-specific deblurring model · Text deblurringdeblurring performanceHurt
use the proposed unpaired method instead of the best reported baselineBMVC_Text · BMVC_TextPSNR: 30.6 -> 22.56Hurt
use the proposed unpaired method instead of the best reported baselineBMVC_Text · BMVC_TextCER: 7.2 -> 10.1Hurt
use the proposed method instead of the blurred-image baselineFace verification · CFP F2Faccuracy: 0.920$\pm$0.014 -> 0.948$\pm$0.006Improved
use the proposed method instead of the blurred-image baselineFace verification · CFP F2Paccuracy: 0.848$\pm$0.013 -> 0.872$\pm$0.015Improved

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