SimCLR learns visual representations by contrasting augmented views of the same image and reaches 76.5% ImageNet top-1 accuracy with a linear classifier, matching a supervised ResNet-50.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2020 2verdicts
ACCEPT 2representative citing papers
Adding an MLP projection head and enhanced augmentations to MoCo produces stronger unsupervised vision baselines that beat SimCLR while using smaller batches.
citing papers explorer
-
A Simple Framework for Contrastive Learning of Visual Representations
SimCLR learns visual representations by contrasting augmented views of the same image and reaches 76.5% ImageNet top-1 accuracy with a linear classifier, matching a supervised ResNet-50.
-
Improved Baselines with Momentum Contrastive Learning
Adding an MLP projection head and enhanced augmentations to MoCo produces stronger unsupervised vision baselines that beat SimCLR while using smaller batches.