Meta-Learning across Meta-Tasks for Few-Shot Learning

Abstract

Existing meta-learning based few-shot learning (FSL) methods typically adopt an episodic training strategy whereby each episode contains a meta-task. Across episodes, these tasks are sampled randomly and their relationships are ignored. In this paper, we argue that the inter-meta-task relationships should be exploited and those tasks are sampled strategically to assist in meta-learning. Specifically, we consider the relationships defined over two types of meta-task pairs and propose different strategies to exploit them. (1) Two meta-tasks with disjoint sets of classes: this pair is interesting because it is reminiscent of the relationship between the source seen classes and target unseen classes, featured with domain gap caused by class differences. A novel learning objective termed meta-domain adaptation (MDA) is proposed to make the meta-learned model more robust to the domain gap. (2) Two meta-tasks with identical sets of classes: this pair is useful because it can be employed to learn models that are robust against poorly sampled few-shots. To that end, a novel meta-knowledge distillation (MKD) objective is formulated. There are some mistakes in the experiments. We thus choose to withdraw this paper.

arXiv commentThere are some mistakes in the experiments. We thus choose to withdraw this paper

References in corpus (14)

• Distilling the Knowledge in a Neural Network
• Prototypical Networks for Few-shot Learning
• Conditional Adversarial Domain Adaptation
• A Simple Neural Attentive Meta-Learner
• Meta-SGD: Learning to Learn Quickly for Few-Shot Learning
• Bridging Theory and Algorithm for Domain Adaptation
• Cross Attention Network for Few-shot Classification
• Cross-Domain Few-Shot Classification via Learned Feature-Wise Transformation
• Infinite Mixture Prototypes for Few-Shot Learning
• TapNet: Neural Network Augmented with Task-Adaptive Projection for Few-Shot Learning
• Domain-Invariant Projection Learning for Zero-Shot Recognition
• Transferring Knowledge across Learning Processes
• Transductive Episodic-Wise Adaptive Metric for Few-Shot Learning
• Learning What and Where to Transfer

Cited by in corpus (3)

• Learning Dynamic Alignment via Meta-filter for Few-shot Learning
• High-order structure preserving graph neural network for few-shot learning
• Few-shot Learning with LSSVM Base Learner and Transductive Modules