SwAV (ResNet-50)

Reported on 3 benchmarks across 3 tasks · 2 papers · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision2 results

Image ClassificationonImageNet - 1% labeled data
Top 5 Accuracy· 2020-06-17
78.5
best: 93.1 (Semi-ViT (ViT-Huge))
SOTA
Unsupervised Learning of Visual Features by Contrasting Cluster Assignments arXiv:2006.09882
Semi-Supervised Image ClassificationonImageNet - 1% labeled data
Top 5 Accuracy· 2020-06-17
78.5
best: 93.1 (Semi-ViT (ViT-Huge))
SOTA
Unsupervised Learning of Visual Features by Contrasting Cluster Assignments arXiv:2006.09882

Methodology2 results

ClassificationonMHIST
Accuracy· uses extra data· 2022-12-09
77.99
best: 88.03 (MoCo-v2 (ResNet-50))
Benchmarking Self-Supervised Learning on Diverse Pathology Datasets arXiv:2212.04690
ClassificationonMHIST
Accuracy· uses extra data
83.21
best: 88.03 (MoCo-v2 (ResNet-50))