RELICv2

Reported on 3 benchmarks across 2 tasks · 1 paper · 2 SOTA

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

Computer Vision3 results

Image ClassificationonImageNet - 1% labeled data
Top 5 Accuracy· 2022-01-13
81.3
best: 93.1 (Semi-ViT (ViT-Huge))
SOTA
Pushing the limits of self-supervised ResNets: Can we outperform supervised learning without labels on ImageNet?arXiv:2201.05119
Semi-Supervised Image ClassificationonImageNet - 1% labeled data
Top 5 Accuracy· 2022-01-13
81.3
best: 93.1 (Semi-ViT (ViT-Huge))
SOTA
Pushing the limits of self-supervised ResNets: Can we outperform supervised learning without labels on ImageNet?arXiv:2201.05119
Image ClassificationonObjectNet
Top-1 Accuracy· uses extra data· 2022-01-13
25.9
best: 82.7 (CoCa)
Pushing the limits of self-supervised ResNets: Can we outperform supervised learning without labels on ImageNet?arXiv:2201.05119