Episodic-free Task Selection for Few-shot Learning

TL;DR

The paper addresses the mismatch between episodic training and testing in few-shot learning by proposing Episodic-free Task Selection (EFTS) that uses episodic tasks only for evaluation. EFTS selects a subset of episodic-free training tasks from a predefined task set based on an inter-task affinity criterion, using episodic tasks to measure transferability via a defined affinity metric. Empirical results on miniImageNet, tiered-ImageNet, and CIFAR-FS show EFTS can match or surpass traditional episodic approaches and non-episodic baselines when evaluated with a nearest-centroid or ProtoNet-style head, highlighting its practical viability. Overall, EFTS demonstrates a scalable, multitask-inspired route to improve few-shot learning without strictly adhering to episodic training.

Abstract

Episodic training is a mainstream training strategy for few-shot learning. In few-shot scenarios, however, this strategy is often inferior to some non-episodic training strategy, e. g., Neighbourhood Component Analysis (NCA), which challenges the principle that training conditions must match testing conditions. Thus, a question is naturally asked: How to search for episodic-free tasks for better few-shot learning? In this work, we propose a novel meta-training framework beyond episodic training. In this framework, episodic tasks are not used directly for training, but for evaluating the effectiveness of some selected episodic-free tasks from a task set that are performed for training the meta-learners. The selection criterion is designed with the affinity, which measures the degree to which loss decreases when executing the target tasks after training with the selected tasks. In experiments, the training task set contains some promising types, e. g., contrastive learning and classification, and the target few-shot tasks are achieved with the nearest centroid classifiers on the miniImageNet, tiered-ImageNet and CIFAR-FS datasets. The experimental results demonstrate the effectiveness of our approach.

Figures (5)

• Figure 1: Test accuracies obtained by using Prototypical Network (ProtoNet), Neighbourhood Component Analysis (NCA) and our proposed Episodic-free Task Selection (EFTS) for 5-way 1-shot (left) and 5-way 5-shot (right) tasks on miniImageNet, respectively. In target tasks, ProtoNet is used as the classifiers.
• Figure 2: The framework of the proposed EFTS. In this task selection section, firstly, using M episodic-free tasks updates the model respectively; Secondly, M updated models are evaluated using the episodic task, and thus M losses are obtained; finally, M losses are compared, and the episodic-free task corresponds to the smallest loss will be selected for the next stage of training. Task selection can be done multiple times during the training process.
• Figure 3: Average affinity against training steps with different Update Number per Affinity (UNA) on CIFAR-FS using ResNet-12. The batchsize is 128 and the training way is 16.
• Figure 4: Average affinity against training steps with different Update Number per Affinity (UNA) on CIFAR-FS using ResNet-12. The batchsize is 128 and the training way is 16.
• Figure 5: Accuracies ($\%$) obtained by EFTS with different interval of task selection against different Q on CIFAR-FS using ResNet-12. The batchsize is 128 and the training way is 16. In addition, NUA = 50 and M = 4.