AI Paper Atlas

Research map / Detection and segmentation / CNN features

FcaNet: Frequency Channel Attention Networks

Zequn Qin, Pengyi Zhang, Fei Wu, Xi Li

ICCV 2021 · Detection and segmentation › CNN features · approach: Efficient network training and design · 1,614 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 global average pooling with multi-spectral DCT compression (LF)SENet / ResNet-50 · ImageNet valTop-1 accuracy: 77.86 -> 78.43Improved
replace GAP with multi-spectral DCT compression (TS)SENet / ResNet-101 · ImageNet valTop-1 accuracy: 79.19 -> 79.63Improved
replace GAP with multi-spectral DCT compression (LF)SENet / ResNet-34 · ImageNet valTop-1 accuracy: 74.83 -> 74.95Improved
replace GAP with multi-spectral DCT compression (LF)SENet / ResNet-152 · ImageNet valTop-1 accuracy: 79.84 -> 80.13Improved
replace GAP with multi-spectral DCT compression (LF)SENet / Faster R-CNN ResNet-50 · COCO val 2017AP: 37.7 -> 39.0Improved
replace GAP with multi-spectral DCT compression (LF)SENet / Faster R-CNN ResNet-101 · COCO val 2017AP: 39.6 -> 41.3Improved
replace GAP with multi-spectral DCT compression (LF)SENet / Mask R-CNN ResNet-50 · COCO val 2017 detectionAP: 38.7 -> 40.3Improved
replace GAP with multi-spectral DCT compression (LF)SENet / Mask R-CNN ResNet-50 · COCO val 2017 instance segmentationAP: 35.4 -> 36.3Improved
select components by two-step top-k instead of lowest frequenciesFcaNet / ResNet-50 · ImageNet valTop-1 accuracy: 78.43 -> 78.57Improved
select components by neural architecture search instead of lowest frequenciesFcaNet / ResNet-50 · ImageNet valTop-1 accuracy: 78.43 -> 78.46No clear effect
use multiple frequency components instead of only GAP (ablation)SENet / ResNet-50 · ImageNet valTop-1 accuracyImproved
use a single non-lowest frequency component instead of GAP (ablation)SENet / ResNet-50 · ImageNet valTop-1 accuracy: <=0.5% Top-1 gapNo clear effect

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