AI Paper Atlas

Research map / Image generation / Image restoration

Rawformer: Unpaired Raw-to-Raw Translation for Learnable Camera ISPs

Georgy Perevozchikov, Nancy Mehta, Mahmoud Afifi, Radu Timofte

ECCV 2024 · Image generation › Image restoration · approach: Image harmonization and color transformation · 11 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 self-supervised inpainting pre-training of the generator (ablation)Rawformer · NUS Canon-to-NikonPSNR: 38.73 -> 41.89Improved
replace L1 loss with L1+VGG then pixel-wise (L1+SSIM+VGG) loss (ablation)Rawformer · NUS Canon-to-NikonPSNR: 37.13 -> 37.94 -> 38.73Improved
add condensed query attention (CQA) to baseline UNet generator (ablation)Rawformer generator · NUS Canon-to-NikonPSNR: 37.23 -> 38.21Improved
add style modulator (ablation)Rawformer generator · NUS Canon-to-NikonPSNR: 38.21 -> 39.08Improved
add scale perceptive feed-forward network (SPFN) (ablation)Rawformer generator · NUS Canon-to-NikonPSNR: 39.08 -> 39.72Improved
add composite up/downsampler (CUp-CDown) (ablation)Rawformer generator · NUS Canon-to-NikonPSNR: 39.72 -> 39.93Improved
add discriminator head with cache on top of attention and batchnorm (ablation)Rawformer discriminator · NUS Canon-to-NikonPSNR: 40.24 -> 41.89Improved
compress queries with average pooling plus linear projection instead of other projections (ablation)CQA block · NUS Canon-to-NikonPSNR: 38.83 (patch merging) -> 41.89Improved
use multi-scale instead of single-scale SPFN (ablation)Rawformer · NUSPSNR: 41.11 -> 41.89Improved
int8 post-training quantization for mobile deploymentRawformer · Raw-to-Raw Samsung-to-iPhonePSNR: 40.98 -> 37.21 (CPU int8, 526 -> 179 ms)Hurt
convert to fp16 with TensorRTRawformer · Raw-to-Raw Samsung-to-iPhonePSNR / time: 40.98 -> 40.92; 26 -> 18 msNo clear effect

Later papers that compared with this method

1 later accepted paper reports Rawformer as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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