On Hard Episodes in Meta-Learning

Samyadeep Basu, Amr Sharaf, Nicolo Fusi, Soheil Feizi

Keywords: meta-learning, few-shot learning

Abstract: Existing meta-learners primarily focus on improving the average task accuracy across multiple episodes. Different episodes, however, may vary in hardness and quality leading to a wide gap in the meta-learner's performance across episodes. Understanding this issue is particularly critical in industrial few-shot settings, where there is limited control over test episodes as they are typically uploaded by end-users. In this paper, we empirically analyse the behaviour of meta-learners on episodes of varying hardness across three standard benchmark datasets: CIFAR-FS, mini-ImageNet, and tiered-ImageNet. Surprisingly, we observe a wide gap in accuracy of around $50\%$ between the hardest and easiest episodes across all the standard benchmarks and meta-learners. We additionally investigate various properties of hard episodes and highlight their connection to catastrophic forgetting during meta-training. To address the issue of sub-par performance on hard episodes, we investigate and benchmark different meta-training strategies based on adversarial training and curriculum learning. We find that adversarial training strategies are much more powerful than curriculum learning in improving the prediction performance on hard episodes.

One-sentence Summary: Analysis of hard episodes in meta-learning