AI Paper Atlas

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

Directional Self-Supervised Learning for Heavy Image Augmentations

Yalong Bai, Yifan Yang, Wei Zhang, Tao Mei

CVPR 2022 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Invariance-based self-supervised learning · 20 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
Add DSSL with heavy augmentationsSimSiam · CIFAR-10 linear evaluationtop-1 accuracy (%): 92.1 -> 94.5Improved
Add DSSL with heavy augmentationsBYOL · ImageNet linear evaluationtop-1 accuracy (%): 67.8 -> 68.3Improved
Add DSSL with heavy augmentationsSimSiam · ImageNet linear evaluationtop-1 accuracy (%): 67.4 -> 68.6Improved
Use RA(8,16) as the heavy augmentation with DSSLBYOL · CIFAR-10 linear evaluationtop-1 accuracy (%): collapse -> 94.0Improved
Combine RA, UA, and Jigsaw as heavy augmentations with DSSLBYOL · ImageNet linear evaluationtop-1 accuracy (%): 67.8 -> 69.3Improved
Add DSSL during pretrainingBYOL · COCO 2017 val object detectionAP: 36.9 -> 37.6Improved
Add DSSL during pretrainingSimSiam · COCO 2017 val instance segmentationAPm: 34.7 -> 35.1Improved
Add symmetric loss among heavily augmented views (ablation)SimSiam with heavy augmentations · CIFAR-10 linear evaluationtop-1 accuracy (%): collapseHurt
Map standard-view features to heavily augmented-view targets (ablation)SimSiam with heavy augmentations · CIFAR-10 linear evaluationtop-1 accuracy (%): 94.20 -> 59.74Hurt
Use directional prediction from heavily augmented views to their source standard views (ablation)SimSiam with heavy augmentations · CIFAR-10 linear evaluationtop-1 accuracy (%): 59.74 -> 93.10Improved
Construct partially ordered views so each hard view is derived from a standard view (ablation)SimSiam with heavy augmentations · CIFAR-10 linear evaluationtop-1 accuracy (%): 93.10 -> 94.20Improved

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