AI Paper Atlas

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

Self-Supervised Learning of Pretext-Invariant Representations

Ishan Misra, Laurens van der Maaten

CVPR 2020 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Invariance-based self-supervised learning · 1,405 citations (OpenAlex)

Check 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
Learn representations invariant to Jigsaw transformations rather than predicting patch permutationsJigsaw self-supervised learning · VOC07 test (VOC07+12 train)APall: 48.9 -> 54.0Improved
Learn representations invariant to Jigsaw transformations rather than predicting patch permutationsJigsaw self-supervised learning · ImageNettop-1 accuracy: 45.7 -> 63.6Improved
Use PIRL instead of supervised pretrainingResNet-50 Faster R-CNN · VOC07 test (VOC07+12 train)APall: 52.6 -> 54.0Improved
Set the PIRL loss mixture to λ=0.5 instead of λ=0 (NPID++) (ablation)PIRL · ImageNettop-1 accuracyImproved
Vary λ away from 0.5 (ablation)PIRL · ImageNettop-1 accuracyMixed
Increase the number of Jigsaw patch permutations (ablation)PIRL · VOC07mAPImproved
Increase the number of negative samples (ablation)PIRL · ImageNettop-1 accuracyImproved
Use rotation-invariant PIRL rather than the covariant Rotation taskRotation pretext learning · ImageNettop-1 accuracy: +11.3Improved
Combine rotation with Jigsaw transforms (ablation)PIRL (Jigsaw) · ImageNettop-1 accuracy: 62.2 -> 63.1Improved
Pretrain on YFCC-1M rather than ImageNetPIRL · ImageNettop-1 accuracy: 63.6 -> 57.8Hurt
Pretrain on YFCC-1M rather than ImageNetPIRL · Places205top-1 accuracy: 49.8 -> 51.0Improved
Double the number of ResNet-50 channels (ablation)PIRL · ImageNettop-1 accuracy: 63.6 -> 67.4Improved

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