AI Paper Atlas

Research map / Image generation / Super-resolution

Face Super-resolution Guided by Facial Component Heatmaps

Xin Yu, Basura Fernando, Bernard Ghanem, Fatih Porikli, Richard Hartley

ECCV 2018 · Image generation › Super-resolution · approach: CNN and transformer SR architectures · 247 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 skip connection (ablation)multi-task upsampling network · Face SR test set (16x16 LR, 8x)PSNR (dB): 22.69 -> 23.14Improved
add autoencoder branch instead of upsampling LR directly (ablation)multi-task upsampling network · Face SR test set (16x16 LR, 8x)PSNR (dB): 22.83 -> 23.14Improved
estimate heatmaps on larger feature maps (16 -> 64) (ablation)HEB · Face SR test set (HEB position ablation)PSNR (dB): 21.97 -> 23.14Improved
estimate heatmaps on 32 instead of 16 feature size (ablation)HEB · Face SR test set (HEB position ablation)PSNR (dB): 21.97 -> 21.98No clear effect
increase hourglass stacks from 1 to 4 (ablation)HEB · Face SR test set (HEB depth ablation)PSNR (dB): 22.32 -> 23.14Improved
increase hourglass stacks from 3 to 4 (ablation)HEB · Face SR test set (HEB depth ablation)PSNR (dB): 22.93 -> 23.14Improved
add heatmap (face structure) loss Lh to pixel loss (ablation)MTUN · Face SR test set (loss ablation)PSNR (dB): 21.43 -> 23.23Improved
add feature-wise loss to pixel + heatmap loss (ablation)MTUN · Face SR test set (loss ablation)PSNR (dB): 23.23 -> 23.35Improved
add discriminative loss to pixel + feature + heatmap loss (ablation)MTUN · Face SR test set (loss ablation)PSNR (dB): 23.35 -> 23.14Mixed
retrain TDAE on the paper's training dataTDAE · Face SR test set (16x16 LR, 8x)PSNR (dB): 18.87 -> 21.39Improved

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