AI Paper Atlas

Research map / Graph learning / GNN architectures

Deep Expander Networks: Efficient Deep Networks from Graph Theory

Ameya Prabhu, Girish Varma, Anoop Namboodiri

ECCV 2018 · Graph learning › GNN architectures · approach: Deep graph convolution and simplification · 52 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 1x1 convs with random-expander X-Conv instead of grouped conv (2x compression) (ablation)MobileNet-0.5 · ImageNetaccuracy: 57.45 -> 58.22Improved
replace grouped conv with X-Conv at 4x compression (ablation)MobileNet-0.5 · ImageNetaccuracy: 49.41 -> 54.00Improved
replace grouped conv with X-Conv at 8x compression (ablation)MobileNet-0.5 · ImageNetaccuracy: 45.13 -> 49.23Improved
replace grouped conv with X-Conv at 16x compression (ablation)MobileNet-0.5 · ImageNetaccuracy: 39.03 -> 44.63Improved
replace dense layers with expander X-Conv (2x) in ResNet-34 vs X-ResNet-2-50ResNet · ImageNetaccuracy: 71.66 -> 72.85Improved
apply X-Conv (2x compression) to ResNet-50ResNet-50 · ImageNetaccuracy: 74.46 -> 72.85Mixed
apply X-Conv (2x) to DenseNet-BC-161DenseNet-BC-161 · ImageNetaccuracy: 76.3 -> 74.3Hurt
compress VGG16 with expander layers (X-VGG16-1, 9x fewer params)VGG16 · CIFAR-10accuracy: 94.0 -> 93.4Hurt
compress DenseNet-BC-40-24 with X-Conv (2x)DenseNet-BC-40-24 · CIFAR-10accuracy: 94.79 -> 94.83No clear effect
compress DenseNet-BC-40-24 with X-Conv (2x)DenseNet-BC-40-24 · CIFAR-100accuracy: 76.05 -> 74.37Hurt
make the X-DenseNet deeper (70 vs 58 layers) at similar compression (ablation)X-DenseNet-BC · CIFAR-100accuracy: 78.74 -> 79.56Improved
widen DenseNet-BC-40-60 to growth rate 200 with 30x compression (ablation)X-DenseNet-BC · CIFAR-100accuracy: 81.62 -> 77.29Hurt

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