AI Paper Atlas

Research map / Video / Video segmentation and tracking

Motion-Guided Latent Diffusion for Temporally Consistent Real-world Video Super-resolution

Xi Yang, Chenhang He, Jianqi Ma, Lei Zhang

ECCV 2024 · Video › Video segmentation and tracking · approach: Real-world video super-resolution · 38 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 motion-guided diffusion sampling (MDS) (ablation)MGLD-VSR baseline · VideoLQwarping error (WE): 9.3069 -> 8.9848Improved
add temporal-aware sequence decoder (TSD) (ablation)MGLD-VSR baseline · VideoLQwarping error (WE): 9.3069 -> 8.3679Improved
combine MDS and TSD (ablation)MGLD-VSR · VideoLQwarping error (WE): 9.3069 -> 7.5478Improved
use RAFT instead of SpyNet flow in motion-guided sampling (ablation)MGLD-VSR · VideoLQwarping error (WE): 7.6734 -> 7.5478No clear effect
use large motion guidance strength eta (ablation)MGLD-VSR · VideoLQvisual qualityHurt
add CFW module (encoder feature) to decoder (ablation)temporal-aware decoder · VideoLQwarping error (WE): 9.8670 -> 8.6846Improved
add temporal modeling (1D conv) to decoder (ablation)VAE decoder · VideoLQwarping error (WE): 9.8670 -> 8.9842Improved
add frame difference loss and structure weighted consistency loss (ablation)decoder fine-tuning · VideoLQwarping error (WE): 7.9480 -> 7.5478Improved
use MGLD-VSR instead of RealBasicVSR for perceptual qualityRealBasicVSR · REDS4LPIPS: 0.4190 -> 0.3776Improved
diffusion prior method versus classic VSR on warping errorMGLD-VSR · REDS4 / UDM10 / SPMCS / VideoLQwarping error (WE): REDS4 6.50 (RVRT) vs 10.08 (Ours)Hurt
use MGLD-VSR (MIMO, lighter CFW) instead of StableSR per frameStableSR · inference speedtime: 3.194s -> 1.113sImproved

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