Uniform Sampling over Episode Difficulty

Abstract

Episodic training is a core ingredient of few-shot learning to train models on tasks with limited labelled data. Despite its success, episodic training remains largely understudied, prompting us to ask the question: what is the best way to sample episodes? In this paper, we first propose a method to approximate episode sampling distributions based on their difficulty. Building on this method, we perform an extensive analysis and find that sampling uniformly over episode difficulty outperforms other sampling schemes, including curriculum and easy-/hard-mining. As the proposed sampling method is algorithm agnostic, we can leverage these insights to improve few-shot learning accuracies across many episodic training algorithms. We demonstrate the efficacy of our method across popular few-shot learning datasets, algorithms, network architectures, and protocols.

NeurIPS'21 camera ready

References in corpus (12)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Theoretical Models of Learning to Learn
• Meta-Learning with Implicit Gradients
• Recasting Gradient-Based Meta-Learning as Hierarchical Bayes
• Accelerating Minibatch Stochastic Gradient Descent using Stratified Sampling
• Meta-Learning with Warped Gradient Descent
• learn2learn: A Library for Meta-Learning Research
• Revisiting Meta-Learning as Supervised Learning
• On Episodes, Prototypical Networks, and Few-shot Learning
• Expert Training: Task Hardness Aware Meta-Learning for Few-Shot Classification
• When Do Curricula Work?
• Adaptive Task Sampling for Meta-Learning