Data-Efficient Image Recognition with Contrastive Predictive Coding

ICLR 2020 Anonymous

Human observers can learn to recognize new categories of objects from a handful of examples, yet doing so with machine perception remains an open challenge. We hypothesize that data-efficient recognition is enabled by representations which make the variability in natural signals more predictable, as suggested by recent perceptual evidence... (read more)

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK LEADERBOARD
Self-Supervised Image Classification ImageNet CPC v2 (ResNet-161) Top 1 Accuracy 71.5% # 4
Top 5 Accuracy 90.1% # 4
Number of Params 305M # 1
Self-Supervised Image Classification ImageNet CPC v2 (ResNet-50) Top 1 Accuracy 63.8% # 14
Top 5 Accuracy 85.3% # 9
Number of Params 24M # 1
Semi-Supervised Image Classification ImageNet - 1% labeled data CPC v2 (ResNet-161) Top 5 Accuracy 77.9% # 3
Top 1 Accuracy 52.7% # 1