Semi-Supervised Few-Shot Learning with a Controlled Degree of Task-Adaptive Conditioning

Sung Whan Yoon, Jun Seo, Jaekyun Moon

Keywords: few-shot learning, meta-learning, semi-supervised learning, task-adaptive clustering, task-adaptive projection space

TL;DR: We propose a semi-supervised few-shot learning algorithm with a controlled degree of task-adaptive conditioning by an iterative update of a task-conditioned projection space where the clustering of unlabeled samples takes place.

Abstract: Few-shot learning aims to handle previously unseen tasks using only a small amount of new training data. In preparing (or meta-training) a few-shot learner, however, massive labeled data are necessary. In the real world, unfortunately, labeled data are expensive and/or scarce. In this work, we propose a few-shot learner that can work well under the semi-supervised setting where a large portion of training data is unlabeled. Our method employs explicit task-conditioning in which unlabeled sample clustering for the current task takes place in a new projection space different from the embedding feature space. The conditioned clustering space is linearly constructed so as to quickly close the gap between the class centroids for the current task and the independent per-class reference vectors meta-trained across tasks. In a more general setting, our method introduces a concept of controlling the degree of task-conditioning for meta-learning: the amount of task-conditioning varies with the number of repetitive updates for the clustering space. During each update, the soft labels of the unlabeled samples estimated in the conditioned clustering space are used to update the class averages in the original embedded space, which in turn are used to reconstruct the clustering space. Extensive simulation results based on the miniImageNet and tieredImageNet datasets show state-of-the-art semi-supervised few-shot classification performance of the proposed method. Simulation results also indicate that the proposed task-adaptive clustering shows graceful degradation with a growing number of distractor samples, i.e., unlabeled samples coming from outside the candidate classes.