Momentum Contrast for Unsupervised Visual Representation Learning

CVPR 2020 Kaiming HeHaoqi FanYuxin WuSaining XieRoss Girshick

We present Momentum Contrast (MoCo) for unsupervised visual representation learning. From a perspective on contrastive learning as dictionary look-up, we build a dynamic dictionary with a queue and a moving-averaged encoder... (read more)

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT LEADERBOARD
Self-Supervised Image Classification ImageNet MoCo (ResNet-50 4x) Top 1 Accuracy 68.6% # 20
Number of Params 375M # 3
Self-Supervised Image Classification ImageNet MoCo (ResNet-50 2x) Top 1 Accuracy 65.4% # 24
Number of Params 94M # 8
Self-Supervised Image Classification ImageNet MoCo (ResNet-50) Top 1 Accuracy 60.6% # 33
Number of Params 24M # 11
Top 1 Accuracy (kNN) 47.1% # 4

Methods used in the Paper