AI Paper Atlas

Research map / Image generation / Super-resolution

Accelerating the Super-Resolution Convolutional Neural Network

Chao Dong, Chen Change Loy, Xiaoou Tang

ECCV 2016 · Image generation › Super-resolution · approach: CNN and transformer SR architectures · 3,842 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
replace last conv layer with a deconvolution layer on the LR input (ablation)SRCNN-Ex · Set5 x3PSNR (dB): 32.83 -> 32.95Improved
replace the 5x5 mapping layer with shrinking, 4 narrow 3x3 mapping layers and expanding (ablation)SRCNN-Ex with deconvolution · Set5 x3PSNR (dB): 32.95 -> 33.01Improved
use smaller first-layer filters and fewer filters (Conv(5,56,1)) (ablation)FSRCNN · Set5 x3PSNR (dB): 33.01 -> 33.06Improved
deconvolution on LR input instead of conv on upsampled input (ablation)SRCNN-Ex · Set5 x3speedup: 1x -> 8.7xImproved
add the expanding layer after the mapping layers (ablation)FSRCNN · Set5PSNR (dB): decreases up to 0.3 dB without itImproved
learn diverse deconvolution kernels per feature map instead of forcing identical kernels (ablation)FSRCNN · Set5PSNR (dB): drops at least 0.9 dB when identicalImproved
increase mapping depth from m=2 to m=4 (d=48, s=12) (ablation)FSRCNN · Set5PSNR (dB): 32.87 -> 33.08Improved
increase d and s from (48,12) to (56,16) at m=4 (ablation)FSRCNN · Set5PSNR (dB): 33.08 -> 33.17No clear effect
smallest network FSRCNN (48,12,2) vs SRCNN-ExFSRCNN · Set5PSNR (dB): 32.75 -> 32.87Improved
transfer conv filters from x3 and fine-tune only the deconvolution for x2 (ablation)FSRCNN (56,12,4) · x2 trainingconvergence / PSNRNo clear effect
use single FSRCNN network for x4 instead of SCN's two x2 modelsFSRCNN (56,12,4) · Set5 x4PSNR (dB): 30.82 -> 30.55Hurt
FSRCNN-s (32,5,1) real-time model vs SRCNNFSRCNN-s · Set5 x3PSNR (dB): 32.45 -> 32.55Improved

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