AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Deep Diffeomorphic Transformer Networks

Nicki Skafte Detlefsen, Oren Freifeld, Søren Hauberg

CVPR 2018 · Architectures and efficiency › CNN design · approach: Dense matching and estimation CNNs · 44 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 an affine and CPAB transformation layerCNN · Distorted MNIST RTSaccuracy: 0.992 -> 0.996Improved
Add an affine and CPAB transformation layerCNN · Distorted MNIST CPABaccuracy: 0.989 -> 0.993Improved
Add a spatial transformerCNN · MNISTclassification performance: 0.991 -> 0.996Improved
Add a CPAB transformer layerCNN · Fashion MNISTclassification performance: 0.922 -> 0.917Hurt
Add affine followed by CPAB transformationsCNN · Fashion MNISTclassification performance: 0.922 -> 0.913Hurt
Add a CPAB transformer layerCNN · CIFAR10classification performance: 0.870 -> 0.895Improved
Add a CPAB transformer layerCNN · CIFAR100classification performance: 0.642 -> 0.659Improved
Add affine followed by CPAB transformationsSiamese network · LFW restrictedclassification performance: 0.788 -> 0.893Improved
Add affine followed by CPAB transformationsCNN · CelebAclassification performance: 0.712 -> 0.772Improved
Add affine followed by CPAB transformationsCNN · LFW unrestrictedtest accuracy: 0.8930 ± 0.0028 -> 0.9543 ± 0.0046Improved
Replace non-diffeomorphic affine transformations with diffeomorphic affine transformations (ablation)Affine spatial transformer · LFW restrictedclassification performance: 0.840 -> 0.842No clear effect
Replace ST-CPAB with ST-Affine+CPAB (ablation)CNN · LFW restrictedclassification performance: 0.878 -> 0.893Improved

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