AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Differentiable Dynamic Wirings for Neural Networks

Kun Yuan, Quanquan Li, Shaopeng Guo, Dapeng Chen, Aojun Zhou, Fengwei Yu, Ziwei Liu

ICCV 2021 · Architectures and efficiency › CNN design · approach: Efficient CNN scaling and dynamic inference · 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
replace static wiring with dynamic instance-aware wiringMobileNetV2-1.0 · ImageNet valtop-1 accuracy (%): 72.60 -> 73.54Improved
replace static wiring with dynamic instance-aware wiringResNet101 · ImageNet valtop-1 accuracy (%): 78.29 -> 79.90Improved
replace static wiring with dynamic instance-aware wiringRegNet-X-600M · ImageNet valtop-1 accuracy (%): 74.03 -> 74.68Improved
use dynamic wiring on deeper ResNets (larger search space)ResNet18 to ResNet101 · ImageNet valtop-1 accuracy (%): 1.02% to 1.61%Improved
replace static wiring with dynamic wiring in detector backboneResNet101 backbone, Faster R-CNN FPN · COCO minivalbox AP (%): 38.59 -> 41.32Improved
replace static wiring with dynamic wiring in detector backboneResNet50 backbone, Faster R-CNN FPN · COCO minivalbox AP (%): 36.42 -> 38.12Improved
use dynamic DDW instead of RandWire random static graphs (WS)MobileNetV2-size networks · ImageNet val (small computation regime)top-1 accuracy (%): 72.26 -> 73.52Improved
use DDW instead of InstaNAS-ImgNet-AMobileNetV2 · ImageNet valtop-1 accuracy (%): 71.9 -> 73.5Improved
add learnable static edge weights (alpha) (ablation)ResNet-50 · ImageNet valtop-1 accuracy (%): 76.70 -> 77.00Improved
add dynamic routing on top of static learnable edge weights (ablation)ResNet-50 · ImageNet valtop-1 accuracy (%): 77.00 -> 78.28Improved
add dynamic routing on top of static learnable edge weights (ablation)ResNet-101 · ImageNet valtop-1 accuracy (%): 78.64 -> 79.90Improved
add dynamic routing on top of static learnable edge weights (ablation)MobileNetV2-1.0 · ImageNet valtop-1 accuracy (%): 72.86 -> 73.54Improved

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