AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Scaling the Scattering Transform: Deep Hybrid Networks

Edouard Oyallon, Eugene Belilovsky, Sergey Zagoruyko

ICCV 2017 · Architectures and efficiency › CNN design · approach: Filter design and trained-filter analysis · 172 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 SIFT+Fisher-vector encoding with scattering + shared local encoderFV + FC · ImageNet ILSVRC2012 valtop-1 accuracy: 55.6 -> 57.0Improved
replace SIFT+Fisher-vector (SVM) with scattering + SLEFV + SVM · ImageNet ILSVRC2012 valtop-1 accuracy: 54.3 -> 57.0Improved
use scattering + SLE (6 learned layers) instead of AlexNetAlexNet · ImageNet ILSVRC2012 valtop-1 error: 43.1 vs 43.0No clear effect
replace early ResNet layers with fixed scattering (Resnet-10 hybrid)Resnet-18 · ImageNet ILSVRC2012 valtop-5 accuracy: 88.8 -> 88.6; 11.7M -> 12.8M paramsNo clear effect
add scattering front-end to wide ResNet in small-sample regimeWRN 16-8 · CIFAR-10 (100 training samples)accuracy: 34.7 -> 38.9Improved
add scattering front-end to wide ResNet in small-sample regimeWRN 16-8 · CIFAR-10 (500 training samples)accuracy: 46.5 -> 54.7Improved
add scattering front-end to wide ResNet in small-sample regimeWRN 16-8 · CIFAR-10 (1000 training samples)accuracy: 60.0 -> 62.0Improved
use hybrid scattering network at full CIFAR-10 data instead of end-to-end WRNScat + Resnet · CIFAR-10 fullaccuracy: 93.1 vs 96.0Hurt
add learned SLE layers F1, F2, F3 on top of scattering (ablation)Raw scattering + SVM · Caltech-101accuracy: 62.8 -> 82.9Improved
use scattering hybrid on STL-10 without unlabeled dataScat + WRN 19-8 · STL-10accuracy: 70.1 -> 76.0Improved

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