AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applications

Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan

ICCV 2023 · Architectures and efficiency › Attention and transformers · approach: Efficient and robust vision transformers · 318 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
replace pool mixers with the proposed Conv. Encoder (ablation)EfficientFormer-L1 · ImageNet-1K valtop-1 accuracy (%): 79.2 -> 79.9Improved
replace self-attention with efficient additive attention (ablation)EfficientFormer-L1 · ImageNet-1K valtop-1 accuracy (%): 79.9 -> 79.7 top-1; latency 1.1 -> 1.0 msMixed
use SwiftFormer encoder (additive attention) in all stages (ablation)SwiftFormer-L1 · ImageNet-1K valtop-1 accuracy (%): 79.7 -> 80.9Improved
SwiftFormer-XS versus MobileViT-XS and EdgeViT-XXSSwiftFormer-XS · ImageNet-1K vallatency iPhone 14 (ms): 0.7 vs 1.5 ms (MobileViT-XS); top-1 75.7 vs 74.8Improved
SwiftFormer-L3 versus EfficientFormer-L3SwiftFormer-L3 · ImageNet-1K valtop-1 accuracy (%): 82.4 -> 83.0Improved
SwiftFormer backbone versus EfficientFormer-L1 on COCO and ADE20KSwiftFormer-L1 · COCO val / ADE20K valAP box / mIoU: 37.9 -> 41.2 AP box; 38.9 -> 41.4 mIoUImproved
SwiftFormer-L1 versus fully convolutional MobileNet-v3 on A100 throughputSwiftFormer-L1 · ImageNet-1K valthroughput A100 (img/s): 4469 vs 10351Hurt

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