Self-Supervised Prototypical Transfer Learning for Few-Shot Classification

2 code implementations • 19 Jun 2020 • Carlos Medina, Arnout Devos, Matthias Grossglauser

Building on these insights and on advances in self-supervised learning, we propose a transfer learning approach which constructs a metric embedding that clusters unlabeled prototypical samples and their augmentations closely together.

Classification General Classification +4