Self-Supervised Learning of Pretext-Invariant Representations

CVPR 2020 Ishan MisraLaurens van der Maaten

The goal of self-supervised learning from images is to construct image representations that are semantically meaningful via pretext tasks that do not require semantic annotations for a large training set of images. Many pretext tasks lead to representations that are covariant with image transformations... (read more)

PDF Abstract CVPR 2020 PDF CVPR 2020 Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Self-Supervised Image Classification ImageNet PIRL Top 1 Accuracy 63.6% # 30
Number of Params 24M # 11
Semi-Supervised Image Classification ImageNet - 10% labeled data PIRL (ResNet-50) Top 5 Accuracy 83.8% # 16

Methods used in the Paper