AI Paper Atlas

Research map / Architectures and efficiency / Neural architecture search

DHP: Differentiable Meta Pruning via HyperNetworks

Yawei Li, Shuhang Gu, Kai Zhang, Luc Van Gool, Radu Timofte

ECCV 2020 · Architectures and efficiency › Neural architecture search · approach: Latency-aware architecture search · 137 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 latent vector sharing among correlated layers (lambda 3^-4, 50% FLOPs) (ablation)DHP ResNet-56 · CIFAR10Top-1 error (%): 7.11 -> 6.52Improved
add latent vector sharing (lambda 3^-4, 38% FLOPs) (ablation)DHP ResNet-56 · CIFAR10Top-1 error (%): 7.98 -> 6.86Improved
add latent vector sharing (lambda 2^-4, 50% FLOPs) (ablation)DHP ResNet-56 · CIFAR10Top-1 error (%): 6.85 -> 7.13Hurt
use l1 instead of l2 sparsity regularizer (ablation)DHP ResNet-20 · CIFAR10Top-1 error (%): 8.66 -> 8.46Improved
use l1 instead of l2 sparsity regularizer (ablation)DHP ResNet-164 · CIFAR10Top-1 error (%): 5.18 -> 5.22No clear effect
prune widened MobileNetV1 (alpha 2) with DHP instead of using originalMobileNetV1 · Tiny-ImageNetTop-1 error (%): 52.71 -> 50.75Improved
prune widened MobileNetV2 (alpha 2) with DHP instead of using originalMobileNetV2 · Tiny-ImageNetTop-1 error (%): 44.75 -> 43.82Improved
use DHP-50 instead of uniform scaling MobileNetV1-0.75MobileNetV1 · Tiny-ImageNetTop-1 error (%): 54.22 -> 51.63Improved
use DHP-10 instead of uniform scaling MobileNetV2-0.3MobileNetV2 · Tiny-ImageNetTop-1 error (%): 53.99 -> 52.43Improved
use DHP instead of MetaPruningMobileNetV2 (11% FLOPs) · Tiny-ImageNetTop-1 error (%): 56.72 -> 52.43Improved
use DHP-60 instead of baseline for SRResNetSRResNet · Urban100 x4 SRPSNR (dB): 25.88 -> 25.76Hurt
use DHP-20 instead of baseline for DnCNNDnCNN · BSD68 denoisingPSNR (dB): 24.93 -> 24.84Hurt

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