AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Adaptative Inference Cost With Convolutional Neural Mixture Models

Adria Ruiz, Jakob Verbeek

ICCV 2019 · Architectures and efficiency › CNN design · approach: Efficient CNN scaling and dynamic inference · 18 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 expectation propagation at inference instead of 1 Monte-Carlo sample (ablation)CNMM · CIFAR-100top-1 accuracy: 71.2 -> 74.4Improved
use 30 Monte-Carlo samples instead of expectation propagation (ablation)CNMM · CIFAR-100top-1 accuracy: 74.4 -> 74.6No clear effect
use 5 Monte-Carlo samples instead of expectation propagation (ablation)CNMM · CIFAR-100top-1 accuracy: 74.4 -> 74.4No clear effect
add stochastic sampling of mixture variables during training (vs training with expectations) (ablation)CNMM · CIFAR-100accuracy under pruningImproved
add stochastic mixture over networks (vs deterministic summation with early exits) (ablation)CNMM · CIFAR-100accuracy vs FLOPsImproved
add intermediate classifiers with per-depth losses (ablation)CNMM · CIFAR-10, CIFAR-100, Cityscapesaccuracy at low FLOPsImproved
add intermediate classifiers, final-step accuracy (ablation)CNMM · CIFAR-10, CIFAR-100, Cityscapesfinal accuracyMixed
prune the trained CNMM by removing low-probability networksCNMM · CIFAR-10/100accuracy vs FLOPs: factor two FLOP reductionNo clear effect
prune the trained CNMM by removing low-probability networksCNMM · Cityscapes valmIoU: about 25% pruningNo clear effect
CNMM instead of ESPNetCNMM · Cityscapes valmIoU: +0.5 points mIoU, FLOPs reduced by 45%Improved

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