Unsupervised meta-learning for few-shot learning

作者：

Highlights：

• Unsupervised meta-learning that auto-constructs tasks from unlabeled data.

• Novel data augmentation method using extra data as prior knowledge.

• Some performance results are close to supervised meta-learning.

摘要

•Unsupervised meta-learning that auto-constructs tasks from unlabeled data.•Novel data augmentation method using extra data as prior knowledge.•Some performance results are close to supervised meta-learning.

论文关键词：Unsupervised learning,Meta-learning,Few-shot learning

论文评审过程：Received 24 September 2020, Revised 10 February 2021, Accepted 1 March 2021, Available online 19 March 2021, Version of Record 26 March 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.107951