Transductive Few-shot Learning with Prototype-based Label Propagation by Iterative Graph Refinement

TL;DR

This paper introduces a novel transductive few-shot learning method that iteratively refines prototypes and graphs based on prototype-sample relations, improving accuracy over existing methods.

Contribution

It proposes a prototype-based label propagation approach with dynamic graph construction and prototype label estimation, addressing limitations of previous prototype and graph-based methods.

Findings

Outperforms state-of-the-art in transductive FSL on multiple datasets.

Effective in semi-supervised FSL with unlabeled data.

Improves prototype estimation and graph construction accuracy.

Abstract

Few-shot learning (FSL) is popular due to its ability to adapt to novel classes. Compared with inductive few-shot learning, transductive models typically perform better as they leverage all samples of the query set. The two existing classes of methods, prototype-based and graph-based, have the disadvantages of inaccurate prototype estimation and sub-optimal graph construction with kernel functions, respectively. In this paper, we propose a novel prototype-based label propagation to solve these issues. Specifically, our graph construction is based on the relation between prototypes and samples rather than between samples. As prototypes are being updated, the graph changes. We also estimate the label of each prototype instead of considering a prototype be the class centre. On mini-ImageNet, tiered-ImageNet, CIFAR-FS and CUB datasets, we show the proposed method outperforms other state-of-the-art methods in transductive FSL and semi-supervised FSL when some unlabeled data accompanies the novel few-shot task.