AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Data-Efficient Image Recognition with Contrastive Predictive Coding

Olivier Henaff

ICML 2020 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Empirical contrastive visual learning · 906 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 the CPC v1 implementation with the CPC v2 architecture and training protocolCPC on ImageNet linear classification · ImageNetTop-1 accuracy: 48.7% to 71.5%Improved
Increase depth and width to the ResNet-161 encoder (ablation)CPC v2 · ImageNet linear classificationTop-1 accuracy: +5%Improved
Increase patch size from 64×64 to 80×80 pixels (ablation)CPC v2 · ImageNet linear classificationTop-1 accuracy: +2%Improved
Use layer normalization instead of batch normalization (ablation)CPC v2 · ImageNet linear classificationaccuracy: +2%Improved
Add bottom-up spatial predictions (ablation)CPC v2 · ImageNet linear classificationaccuracy: +2%Improved
Predict in all four spatial directions (ablation)CPC v2 · ImageNet linear classificationaccuracy: +2.5%Improved
Randomly drop two color channels in each patch (ablation)CPC v2 · ImageNet linear classificationaccuracy: +3%Improved
Add generic transformations, elastic deformations, and color transforms to patch augmentation (ablation)CPC v2 · ImageNet linear classificationaccuracy: +4.5%Improved
Train the classifier on CPC features rather than raw pixelsImageNet classifiers with limited labels · ImageNet, 1% labelsTop-5 accuracy: 44.1 -> 78.3Improved
Fine-tune CPC features for the supervised classification task (ablation)CPC v2 classifier · ImageNet, 1% labelsTop-5 accuracy: 77.1 -> 78.3Improved
Pre-train the representation with CPC v2 rather than supervised ImageNet pre-trainingFaster R-CNN object detector · PASCAL VOC 2007mAP: 74.7 -> 76.6Improved

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