Self-Supervised Prototypical Transfer Learning for Few-Shot Classification

TL;DR

This paper introduces ProtoTransfer, a self-supervised transfer learning method that creates a metric embedding for few-shot classification, outperforming unsupervised meta-learning and matching supervised methods with fewer labels.

Contribution

The paper presents a novel self-supervised prototypical transfer learning approach that constructs an effective embedding for few-shot learning, reducing label requirements and handling domain shifts.

Findings

ProtoTransfer outperforms state-of-the-art unsupervised meta-learning on mini-ImageNet.

It achieves comparable performance to supervised methods with significantly fewer labels.

The approach effectively handles domain shifts in few-shot tasks.

Abstract

Most approaches in few-shot learning rely on costly annotated data related to the goal task domain during (pre-)training. Recently, unsupervised meta-learning methods have exchanged the annotation requirement for a reduction in few-shot classification performance. Simultaneously, in settings with realistic domain shift, common transfer learning has been shown to outperform supervised meta-learning. 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. This pre-trained embedding is a starting point for few-shot classification by summarizing class clusters and fine-tuning. We demonstrate that our self-supervised prototypical transfer learning approach ProtoTransfer outperforms state-of-the-art unsupervised meta-learning methods on few-shot tasks from the mini-ImageNet dataset. In few-shot experiments with domain shift, our approach even has comparable performance to supervised methods, but requires orders of magnitude fewer labels.