AI Paper Atlas

Research map / Detection and segmentation / CNN features

ExpandNets: Linear Over-parameterization to Train Compact Convolutional Networks

Shuxuan Guo, Jose M. Alvarez, Mathieu Salzmann

NeurIPS 2020 · Detection and segmentation › CNN features · approach: Efficient network training and design

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
expand convolutional kernels (CK) instead of only FC layers (ablation)SmallNet 7x7 · CIFAR-10Top-1 accuracy: 78.64 -> 80.27Improved
expand conv layers with CL (ablation)SmallNet 7x7 · CIFAR-10Top-1 accuracy: 78.63 -> 78.47No clear effect
add FC expansion to CK (ablation)ExpandNet-CK · CIFAR-100Top-1 accuracy: 48.55 -> 48.62Improved
add knowledge distillation to ExpandNet-CK+FCExpandNet-CK+FC · CIFAR-100Top-1 accuracy: 48.62 -> 49.13Improved
add nonlinear-counterpart initialization (ablation)ExpandNet-CK+FC · CIFAR-100Top-1 accuracy: 48.62 -> 49.82Improved
add nonlinear-counterpart initialization (ablation)YOLO-LITE ExpandNet-CL · PASCAL VOC2007 testmAP: 30.97 -> 35.14Improved
expand all 3x3 convs with CLMobileNet · ImageNet valTop-1 accuracy: 66.48 -> 69.40Improved
expand non-residual 3x3 convs with CLMobileNetV2 · ImageNet valTop-1 accuracy: 63.75 -> 65.62Improved
expand with CLShuffleNetV2 0.5x · ImageNet valTop-1 accuracy: 56.89 -> 57.38Improved
train baseline longer to match ExpandNet training time (ablation)MobileNet · CIFAR-10Top-1 accuracy: 89.61 -> 88.87Hurt
initialize compact net with contracted ExpandNet instead of training ExpandNet (ablation)MobileNetV2 · ImageNet valTop-1 accuracy: 63.75 -> 63.07Hurt
lower expansion rate r from 4 to 0.25 (ablation)ExpandNet-CK · CIFAR-10Top-1 accuracy: 80.27 -> 72.32Hurt

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