AI Paper Atlas

Research map / Architectures and efficiency / CNN design

XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks

Mohammad Rastegari, Vicente Ordonez, Joseph Redmon, Ali Farhadi

ECCV 2016 · Architectures and efficiency › CNN design · approach: Filter design and trained-filter analysis · 4,554 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
compute scaling factor alpha in closed form instead of as a learned layer parameter (ablation)BWN AlexNet · ImageNet valtop-1 accuracy: 46.2 -> 56.8Improved
compute scaling factor alpha in closed form instead of as a learned layer parameter (ablation)BWN AlexNet · ImageNet valtop-5 accuracy: 69.5 -> 79.4Improved
reorder block to B-A-C-P instead of C-B-A-P (ablation)XNOR-Net AlexNet · ImageNet valtop-1 accuracy: 30.3 -> 44.2Improved
reorder block to B-A-C-P instead of C-B-A-P (ablation)XNOR-Net AlexNet · ImageNet valtop-5 accuracy: 57.5 -> 69.2Improved
add input scaling factor beta (ablation)XNOR-Net AlexNet · ImageNet valtop-1 accuracy: less than 1%No clear effect
binarize gradients in backward pass (ablation)XNOR-Net AlexNet · ImageNet valtop-1 accuracy: drop only by 1.4%Hurt
binarize weights (BWN) versus full precisionAlexNet · ImageNet valtop-1 accuracy: 56.6 -> 56.8No clear effect
BWN versus BinaryConnectAlexNet · ImageNet valtop-1 accuracy: 35.4 -> 56.8Improved
XNOR-Net versus BinaryNetAlexNet · ImageNet valtop-1 accuracy: 27.9 -> 44.2Improved
binarize weights (BWN) versus full precisionResNet-18 · ImageNet valtop-1 accuracy: 69.3 -> 60.8Hurt
binarize weights and inputs (XNOR-Net) versus full precisionResNet-18 · ImageNet valtop-1 accuracy: 69.3 -> 51.2Hurt
binarize weights (BWN) versus full precisionGoogLenet variant · ImageNet valtop-1 accuracy: 71.3 -> 65.5Hurt

Later papers that compared with this method

20 later accepted papers report XNOR-Net as a baseline in their results tables; 20 of them report a better number on most of the results they share with it.

12 more not listed.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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