Semi-Supervised Few-Shot Learning with Prototypical Random Walks

TL;DR

This paper introduces Prototypical Random Walk Networks, a semi-supervised few-shot learning method that improves class representations using a novel random walk loss, outperforming existing approaches even with limited labeled data.

Contribution

The authors propose a new semi-supervised few-shot learning approach that models prototypical random walks without extra graph neural network parameters, achieving superior performance and robustness.

Findings

Outperforms baselines on most benchmarks

Achieves 50.89% accuracy in 1-shot mini-ImageNet with 40% labeled data

Demonstrates robustness to distractors and class distribution mismatch

Abstract

Recent progress has shown that few-shot learning can be improved with access to unlabelled data, known as semi-supervised few-shot learning(SS-FSL). We introduce an SS-FSL approach, dubbed as Prototypical Random Walk Networks(PRWN), built on top of Prototypical Networks (PN). We develop a random walk semi-supervised loss that enables the network to learn representations that are compact and well-separated. Our work is related to the very recent development of graph-based approaches for few-shot learning. However, we show that compact and well-separated class representations can be achieved by modeling our prototypical random walk notion without needing additional graph-NN parameters or requiring a transductive setting where a collective test set is provided. Our model outperforms baselines in most benchmarks with significant improvements in some cases. Our model, trained with 40$\%$ of the data as labeled, compares competitively against fully supervised prototypical networks, trained on 100$\%$ of the labels, even outperforming it in the 1-shot mini-Imagenet case with 50.89$\%$ to 49.4$\%$ accuracy. We also show that our loss is resistant to distractors, unlabeled data that does not belong to any of the training classes, and hence reflecting robustness to labeled/unlabeled class distribution mismatch. Associated GitHub page can be found at https://prototypical-random-walk.github.io.