AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Devil is in the Uniformity: Exploring Diverse Learners within Transformer for Image Restoration

Shihao Zhou, Dayu Li, Jinshan Pan, Juncheng Zhou, Jinglei Shi, Jufeng Yang

ICCV 2025 · Architectures and efficiency › Attention and transformers · approach: Efficient image transformers · 13 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
use HMHA instead of W-MSA (ablation)HINT · LOL-v2-synPSNR: 24.19 -> 27.17Improved
use HMHA instead of MDTA (ablation)HINT · LOL-v2-synPSNR: 26.42 -> 27.17Improved
use similarity reranking instead of random shuffle before the split (ablation)HMHA · LOL-v2-synPSNR: 26.54 -> 27.17Improved
add reranking and hierarchical split versus no ranking (ablation)HMHA · LOL-v2-synPSNR: 26.42 -> 27.17Improved
add QKCU (IntraCache and InterCache) (ablation)HINT · LOL-v2-synPSNR: 26.47 -> 27.17Improved
add IntraCache only (ablation)HINT · LOL-v2-synPSNR: 26.47 -> 26.67Improved
add InterCache only (ablation)HINT · LOL-v2-synPSNR: 26.47 -> 26.72Improved
add InterCache on top of IntraCache (ablation)HINT · LOL-v2-synPSNR: 26.67 -> 27.17Improved
add IntraCache on top of InterCache (ablation)HINT · LOL-v2-synPSNR: 26.72 -> 27.17Improved
add QKCU modules (ablation)HINT · LOL-v2-synparameters (M): 21.34 -> 24.87Hurt
add reranking to HMHA (ablation)HMHA · LOL-v2-synparameters (M): 24.87 -> 24.87No clear effect
apply HINT enhancement before detectorYOLO-v3 · ExDarkmean AP: 45.1 -> 52.7Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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