AI Paper Atlas

Research map / Video / Video segmentation and tracking

Real-World Video Super-Resolution: A Benchmark Dataset and a Decomposition Based Learning Scheme

Xi Yang, Wangmeng Xiang, Hui Zeng, Lei Zhang

ICCV 2021 · Video › Video segmentation and tracking · approach: Space-time video super-resolution · 65 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
train on RealVSR instead of synthetic Vimeo-90kEDVR · RealVSR test setPSNR (Y): 24.58 -> 25.60Improved
train on RealVSR instead of synthetic Vimeo-90kRCAN · RealVSR test setPSNR (Y): 24.66 -> 25.50Improved
train on Vimeo-90k instead of nothing (bicubic)FSTRN · RealVSR test setPSNR (Y): 24.67 -> 24.63Hurt
train on RealVSR instead of Vimeo-90kEDVR · RealVSR test setSSIM (Y): 0.7838 -> 0.8102Improved
train on RealVSR instead of Vimeo-90kEDVR · RealVSR test setNIQE: 4.6377 -> 3.9082Improved
train on RealVSR instead of Vimeo-90kEDVR · RealVSR test setT-diff: 4.4229 -> 3.6860Improved
add Laplacian pyramid losses Ls and Ld on Y channel (ablation)EDVR · RealVSR test setPSNR (Y): 25.60 -> 25.83Improved
add Laplacian pyramid losses Ls and Ld on Y channel (ablation)TOF · RealVSR test setPSNR (Y): 25.59 -> 25.65Improved
add gradient weighted chroma loss Lc (ablation)EDVR · RealVSR test setPSNR (Y): 25.83 -> 25.83No clear effect
replace RaGAN discriminator with multi-scale edge-based GAN (Lv2) (ablation)EDVR · RealVSR test setLPIPS: 0.2598 -> 0.2459Improved
replace RaGAN discriminator with multi-scale edge-based GAN (Lv2) (ablation)EDVR · RealVSR test setDISTS: 0.0852 -> 0.0766Improved
replace RaGAN discriminator with multi-scale edge-based GAN (Lv2) (ablation)TOF · RealVSR test setLPIPS: 0.2625 -> 0.2622Improved

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