AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Channel Equilibrium Networks for Learning Deep Representation

Wenqi Shao, Shitao Tang, Xingang Pan, Ping Tan, Xiaogang Wang, Ping Luo

ICML 2020 · Architectures and efficiency › CNN design · approach: Compact CNN architectures · 5 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 CE blockResNet50 · ImageNet valtop-1 accuracy: 76.6 -> 78.3Improved
Add CE blockResNet18 · ImageNet valtop-1 accuracy: 70.4 -> 71.9Improved
Add CE blockResNet101 · ImageNet valtop-1 accuracy: 78.0 -> 79.0Improved
Add CE blockMobileNetv2 1× · ImageNet valtop-1 accuracy: 72.5 -> 74.6Improved
Add CE blockShuffleNetv2 0.5× · ImageNet valtop-1 accuracy: 59.2 -> 60.5Improved
Add CE blockShuffleNetv2 1× · ImageNet valtop-1 accuracy: 69.0 -> 71.2Improved
Add CE blockMask R-CNN ResNet50 · COCO2017 mini-valAP^b: 38.6 -> 40.8Improved
Add CE blockMask R-CNN ResNet50 · COCO2017 mini-valAP^m: 34.2 -> 36.9Improved
Use synchronized CE statisticsMask R-CNN ResNet50 · COCO2017 mini-valAP^b: 40.8 -> 42.0Improved
Use the BD branch without IR (ablation)ResNet50 · ImageNet valtop-1 accuracy: 76.6 -> 77.0Improved
Use the IR branch without BD (ablation)ResNet50 · ImageNet valtop-1 accuracy: 76.6 -> 77.3Improved
Combine BD and IR branches (ablation)ResNet50 · ImageNet valtop-1 accuracy: 76.6 -> 78.3 (+1.7)Improved

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