AI Paper Atlas

Research map / 3D vision / Face and shape

Motion Deblurring of Faces

Grigorios G. Chrysos, Paolo Favaro, Stefanos P. Zafeiriou

IJCV 2018 · 3D vision › Face and shape · approach: 3D morphable face model regression · 31 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 HG conditioning to the blur-removal generator (ablation)face deblurring GAN · 300VW (PrAvg 7)AUC: 0.759 -> 0.783Improved
add HG conditioning to the generator (ablation)face deblurring GAN · 300VW (PrAvg 7)failure rate (%): 7.895 -> 5.962Improved
add additional label data (2MF2 backward data) to training (ablation)identity-noHG · 300VW (PrAvg 7)AUC: 0.751 -> 0.759Improved
add the identity branch (ablation)face deblurring GAN · 300VW (PrAvg 7)AUC: 0.783 -> 0.798Improved
add the identity branch (ablation)face deblurring GAN · 300VW (PrAvg 7)failure rate (%): 5.962 -> 3.947Improved
add the identity branch (ablation)face deblurring GAN · 300VW (PrAvg 7)SSIM: 0.92883 -> 0.93501Improved
deblur with the proposed method before landmark localization instead of leaving blurredlandmark localization · 300VW (PrAvg 7)AUC: 0.743 -> 0.798Improved
use ours instead of Nah et al (2017)landmark localization · 300VW (PrAvg 7)AUC: 0.792 -> 0.798Improved
use ours instead of Nah et al (2017) under non-averaging VLA blurlandmark localization · 300VW (VLA)AUC: 0.767 -> 0.786Improved
increase averaged frames from 7 to 21proposed method · 300VWAUC: 0.798 -> 0.748Hurt
use optimization-based deblurring (Xu et al, Chakrabarti) instead of learning-basedlandmark localization · 300VW (PrAvg 15)AUC: 0.446 and 0.437 vs 0.769Hurt
deblur with Nah et al (2017) before localization instead of blurred inputlandmark localization · 300VW (PrAvg 21)failure rate (%): 9.270 -> 7.255Improved

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