AI Paper Atlas

Research map / Image generation / Super-resolution

EfficientViT: Lightweight Multi-Scale Attention for High-Resolution Dense Prediction

Han Cai, Junyan Li, Muyan Hu, Chuang Gan, Song Han

ICCV 2023 · Image generation › Super-resolution · approach: Perceptual and real-world SR · 396 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 both multi-scale learning and global attention to the module (ablation)EfficientViT module · Cityscapes (from scratch, equal MACs)mIoU: 68.1 -> 74.5Improved
add one of the two components (multi-scale or global attention) alone (ablation)EfficientViT module · Cityscapes (from scratch, equal MACs)mIoU: 68.1 -> 72.3 and 68.1 -> 72.2Improved
replace SegFormer-B1 with EfficientViT-B1semantic segmentation model · Cityscapes valmIoU: 78.5 -> 80.5Improved
replace SegNeXt-T with EfficientViT-B1semantic segmentation model · Cityscapes valOrin latency: 93.2ms -> 24.3msImproved
replace SegFormer-B1 with EfficientViT-B1semantic segmentation model · ADE20K valmIoU: 42.2 -> 42.8Improved
replace SegNeXt-S with EfficientViT-B2semantic segmentation model · ADE20K valmIoU: 44.3 -> 45.9Improved
use EfficientViT-B3 instead of EfficientNet-B6 for classificationimage classifier · ImageNet-1k valtop-1 accuracy: 84.0 -> 84.2Improved
replace Restormer with EfficientViT for super-resolutionsuper-resolution model · FFHQ super-resolution (512x512 to 1024x1024)PSNR: 43.43 -> 43.58Improved
replace Restormer with EfficientViT for lightweight super-resolutionsuper-resolution model · BSD100 super-resolution (160x240 to 320x480)PSNR: 32.31 -> 32.33No clear effect
replace softmax attention with ReLU-based linear attention (ablation)attention module · mobile CPU latency (Snapdragon 855)latency speedup: 3.3-4.5x fasterImproved

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