AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Cross-Convolutional-Layer Pooling for Image Recognition

Lingqiao Liu, Chunhua Shen, Anton van den Hengel

TPAMI 2016 · Architectures and efficiency › CNN design · approach: Filter design and trained-filter analysis · 95 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
use cross-layer pooling instead of SCFV on OConv, AlexNetSCFV · MIT-67accuracy: 59.2 -> 63.0Improved
use cross-layer pooling instead of SCFV on OConv, VGGNetSCFV · Birds-200accuracy: 73.7 -> 77.0Improved
use AConv layer instead of OConv for cross-layer pooling on Birds-200 (ablation)CrossLayer VGGNet · Birds-200accuracy: 77.0 -> 69.4Hurt
use AConv layer instead of OConv for cross-layer pooling on PASCAL-07 (ablation)CrossLayer VGGNet · PASCAL VOC 2007mAP: 84.1 -> 87.1Improved
add PCA to the full normalization pipeline (ablation)CrossLayer AConv MIT-67 · MIT-67accuracy: 72.6 -> 78.2Improved
add power normalization (ablation)CrossLayer AConv MIT-67 · MIT-67accuracy: 76.4 -> 78.2Improved
add l2 normalization (ablation)CrossLayer AConv MIT-67 · MIT-67accuracy: 77.1 -> 78.2Improved
use deeper layers conv5-3/conv5-4 instead of conv5-2/conv5-3 (ablation)CrossLayer VGGNet · PASCAL VOC 2007mAP: 82.2 -> 84.1Improved
use sum-pooling instead of max-pooling for channel pooling (ablation)CrossLayer VGGNet · MIT-67accuracy: 72.8 -> 74.4Improved
use max-pooling instead of sum-pooling for channel pooling (ablation)CrossLayer VGGNet · PASCAL VOC 2007mAP: 84.1 -> 85.2Mixed
apply feature sign quantization (2 bits per dimension) (ablation)CrossLayer VGGNet · PASCAL VOC 2007mAP: 87.0 -> 85.1Hurt
apply feature sign quantization (2 bits per dimension) (ablation)CrossLayer VGGNet · MIT-67accuracy: 78.2 -> 77.9No 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