AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Deep CNNs Meet Global Covariance Pooling: Better Representation and Generalization

Qilong Wang, Jiangtao Xie, Wangmeng Zuo, Lei Zhang, Peihua Li

TPAMI 2020 · Architectures and efficiency › CNN design · approach: Higher-order pooling · 113 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 matrix power normalization (MPN) to covariance pooling (ablation)plain COV · ImageNet val, AlexNet, 1-crop (normalization ablation)top-1 error (%): 40.41 -> 38.51Improved
add M-Fro normalization after MPN (ablation)MPN-COV · ImageNet val, AlexNet, 1-crop (normalization ablation)top-1 error (%): 38.51 -> 39.93Hurt
add M-l2 normalization after MPN (ablation)MPN-COV · ImageNet val, AlexNet, 1-crop (normalization ablation)top-1 error (%): 38.51 -> 39.62Hurt
add E-PN after MPN (ablation)MPN-COV · ImageNet val, AlexNet, 1-crop (normalization ablation)top-1 error (%): 38.51 -> 40.75Hurt
add E-PN after iSQRT-COV (ablation)iSQRT-COV-Net · Places365 val, 10-croptop-1 error (%): 43.68 -> 45.34Hurt
add trace post-compensation instead of BN (ablation)iSQRT-COV-Net · ImageNet val, ResNet-50, 1-crop (post-compensation)top-1 error (%): 23.12 -> 22.14Improved
add post-compensation (versus none) (ablation)iSQRT-COV-Net · ImageNet val, ResNet-50, 1-crop (post-compensation)top-1 error (%): fails to converge without itImproved
use progressive dimensionality reduction (2048->512->128 instead of 2048->128) (ablation)iSQRT-COV-Net · ImageNet val, ResNet-50, 1-crop (compact representation)top-1 error (%): 22.78 -> 22.33Improved
use progressive dimensionality reduction (2048->256->64 instead of 2048->64) (ablation)iSQRT-COV-Net · ImageNet val, ResNet-50, 1-crop (compact representation)top-1 error (%): 23.73 -> 22.98Improved
reduce covariance dimension from 256 to 64 with a single 1x1 conv (ablation)iSQRT-COV-Net · ImageNet val, ResNet-50, 1-crop (compact representation)top-1 error (%): 22.14 -> 23.73Hurt
use trace instead of Frobenius pre-normalization (ablation)iSQRT-COV-Net · ImageNet val, AlexNet, 1-crop (covariance pooling comparison)top-1 error (%): 38.78 -> 38.45Improved
add first-order mean via Gaussian embedding (G2DeNet vs MPN-COV)MPN-COV-Net · ImageNet val, AlexNet, 1-crop (covariance pooling comparison)top-1 error (%): 38.51 -> 38.48No clear effect

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