AI Paper Atlas

Research map / Image generation / Super-resolution

HiT-SR: Hierarchical Transformer for Efficient Image Super-Resolution

XIANG ZHANG, Yulun Zhang, Fisher Yu

ECCV 2024 · Image generation › Super-resolution · approach: Efficient lightweight super-resolution · 83 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
convert SwinIR-Light to hierarchical transformer with SCC (HiT-SIR)SwinIR-Light · Urban100 x2PSNR: 32.76 -> 33.02Improved
replace shifted window self-attention with SCCSwinIR-Light · A100 runtime, x2, 720x1280 outputruntime (ms): 1899 -> 249Improved
convert SRFormer-Light to HiT-SRFSRFormer-Light · A100 runtime, x2, 720x1280 outputruntime (ms): 2593 -> 331Improved
use expanding hierarchical windows instead of fixed small windows (ablation)HiT-SIR · Manga109 x2PSNR: 38.88 -> 39.21Improved
use hierarchical windows instead of fixed large 64x64 windows (ablation)HiT-SIR · Manga109 x2PSNR: 39.09 -> 39.21; FLOPs 241.4G -> 209.9GImproved
expand windows small to large rather than shrink large to small (ablation)HiT-SIR · Manga109 x2PSNR: 39.12 -> 39.21Improved
use dual feature extraction (linear + conv branches) instead of linear projection (ablation)HiT-SIR · Manga109 x2PSNR: 38.99 -> 39.21Improved
use queries and values only, estimating no separate keys (ablation)HiT-SIR · Manga109 x2PSNR: 39.19 -> 39.21; params 826K -> 772KNo clear effect
use correlation instead of softmax attention (ablation)HiT-SIR · Manga109 x2PSNR: 39.08 -> 39.21Improved
combine spatial and channel self-correlation rather than either alone (ablation)HiT-SIR · Manga109 x2PSNR: 39.04 (S-SC only), 39.18 (C-SC only) -> 39.21Improved
multi-head for S-SC and single-head for C-SC instead of uniform head strategy (ablation)HiT-SIR · Manga109 x2PSNR: 39.14 (all single) and 39.08 (all multi) -> 39.21Improved
HiT-SIR with hierarchical windows trains fasterSwinIR-Light · Urban100 x2, Manga109 x2iterations to match performanceImproved

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