AI Paper Atlas

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

SCAN: Learning to Classify Images without Labels

Wouter Van Gansbeke, Simon Vandenhende, Stamatios Georgoulis, Marc Proesmans, Luc Van Gool

ECCV 2020 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Pretext-task visual pre-training · 579 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 K-means with SCAN-loss clustering (SimCLR augmentation) (ablation)Pretext features · CIFAR10accuracy: 65.9 -> 78.7Improved
use RandAugment instead of SimCLR augmentations in SCAN-loss (ablation)SCAN-loss · CIFAR10accuracy: 78.7 -> 81.8Improved
add self-labeling after clustering (ablation)SCAN · CIFAR10accuracy: 81.8 -> 87.6Improved
use strong augmentation (RA) rather than SimCLR augmentation in self-labeling (ablation)SCAN self-labeling · CIFAR10accuracy: 10.0 -> 87.4Improved
use instance discrimination pretext instead of RotNet (ablation)SCAN · CIFAR10accuracy: 74.3 -> 87.6Improved
use instance discrimination (SimCLR) pretext instead of RotNet with K-means (ablation)K-means · CIFAR10accuracy: 27.1 -> 65.9Improved
use nearest neighbors K=5 instead of K=0 (ablation)SCAN · CIFAR10accuracy: 56.3 -> 79.3Improved
use nearest neighbors K=5 instead of K=0 (ablation)SCAN · CIFAR100-20accuracy: 24.6 -> 41.1Improved
use nearest neighbors K=5 instead of K=0 (ablation)SCAN · STL10accuracy: 47.70 -> 69.8Improved
overcluster to 2x the number of classesSCAN · CIFAR100-20accuracy: 45.9 -> 55.1Improved
overcluster to 2x the number of classesSCAN · CIFAR10accuracy: 87.6 -> 86.2No clear effect
overcluster to 2x the number of classesSCAN · STL10accuracy: 76.7 -> 76.8No clear effect

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