AI Paper Atlas

Research map / Image generation / Image restoration

SQAD: Automatic Smartphone Camera Quality Assessment and Benchmarking

Zilin Fang, Andrey Ignatov, Eduard Zamfir, Radu Timofte

ICCV 2023 · Image generation › Image restoration · approach: Raw and low-light noise modeling · 6 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
use MANIQA instead of ResNet50camera device classifier · SQAD testTop-1 accuracy: 73.33 -> 93.14Improved
resize test images to ratio 0.7 then 0.5 (ablation)ResNet50 classifier · SQAD testTop-1 accuracy: 49.4 -> 15.7Hurt
use multi-crop (16x) predictions at test time (ablation)quality factor regressors · SQAD testSROCC/PLCCImproved
increase number of test crops (ablation)ResNet50 · SQAD testSROCC/PLCCImproved
use more test samples per device (ablation)ResNet50 · SQAD testSROCC/PLCCImproved
use small training crop sizes (ablation)ResNet50 · SQAD testSROCC/PLCCHurt
replace ResNet50 regressor with SVR on its features (ablation)cross-camera resolution · SQAD held-out devicesSROCC: .6212 -> .6415 SROCC; .6075 -> .5141 PLCCMixed
normalize quality factors to [0,1] before regression (ablation)quality regressors · SQADconvergenceImproved

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