atlas

Research map / Architectures and efficiency / CNN design

STAR-Caps: Capsule Networks with Straight-Through Attentive Routing

NeurIPS 2019 · Architectures and efficiency › CNN design · approach: Efficient CNN scaling and dynamic inference · 44 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (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 EM routing with straight-through attentive routingEMCaps:{32,4,64,4,10} · MNISTaccuracy: 96.89 -> 99.49Helped
replace EM routing with straight-through attentive routingEMCaps:{64,8,64,8,10} · MNISTaccuracy: 98.12 -> 99.57Helped
replace EM routing with straight-through attentive routingEMCaps · SmallNorbaccuracy: 97.8% -> 98.0%; 98.2% -> 98.2%No clear effect
replace EM routing with straight-through attentive routingEMCaps · SmallNorb novel viewpointsaccuracy: 86.5 -> 86.3No clear effect
replace EM routing with straight-through attentive routingEMCaps · CIFAR-10accuracy: 88.10% -> 91.23%Helped
replace EM routing with straight-through attentive routingEMCaps · CIFAR-100accuracy: 19% -> 67.66%Helped
use capsule network instead of CNN on novel viewpointsSTAR-CAPS · SmallNorb novel viewpointsaccuracy: 80.0 -> 86.3Helped

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