AI Paper Atlas

Research map / Detection and segmentation / CNN features

An Image Patch Is a Wave: Phase-Aware Vision MLP

Yehui Tang, Kai Han, Jianyuan Guo, Chang Xu, Yanxi Li, Chao Xu, Yunhe Wang

CVPR 2022 · Detection and segmentation › CNN features · approach: Plain and high-resolution ViTs · 144 citations (OpenAlex)

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 fixed real-valued token mixing with input-dependent dynamic phase modulation (ablation)Wave-MLP-T* · ImageNetTop-1 accuracy: 78.8 -> 80.1Improved
Add input-dependent dynamic phase rather than static learned phase (ablation)Wave-MLP-T* · ImageNetTop-1 accuracy: 79.3 -> 80.1Improved
Use depth-wise convolution rather than identity projection to estimate phase (ablation)Wave-MLP-T* · ImageNetTop-1 accuracy: 79.3 -> 80.1Improved
Use channel-FC rather than depth-wise convolution to estimate phase (ablation)Wave-MLP-T* · ImageNetTop-1 accuracy: 80.1 -> 80.6Improved
Increase local token-mixing window size from 3 to 7 (ablation)Wave-MLP-T* · ImageNetTop-1 accuracy: 79.7 -> 80.1Improved
Use all-token mixing instead of window size 7 (ablation)Wave-MLP-T* · ImageNetTop-1 accuracy: 80.1 -> 80.0No clear effect
Use Wave-MLP-S instead of Swin-T as the ImageNet classifierImageNet classification · ImageNet valTop-1 accuracy: 81.3 -> 82.6Improved
Use Wave-MLP-S instead of Swin-T as the RetinaNet backboneRetinaNet 1× · COCO val2017AP: 41.5 -> 43.4Improved
Use Wave-MLP-S instead of Swin-T as the Mask R-CNN backboneMask R-CNN 1× · COCO val2017Box AP: 42.2 -> 44.0Improved
Use Wave-MLP-S instead of Swin-T as the Mask R-CNN backboneMask R-CNN 1× · COCO val2017Mask AP: 39.1 -> 40.0Improved
Use Wave-MLP-M instead of CycleMLP-B3 as the Semantic FPN backboneSemantic FPN · ADE20K valmIoU: 44.5 -> 46.8Improved

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