Transductive Information Maximization For Few-Shot Learning

TL;DR

This paper presents Transductive Information Maximization (TIM), a novel method for few-shot learning that enhances accuracy by maximizing mutual information between query features and labels, with a faster inference solver and broad applicability.

Contribution

TIM introduces a mutual information maximization approach with an efficient solver, improving transductive few-shot learning performance without complex meta-learning.

Findings

Outperforms state-of-the-art methods across datasets.

Achieves 2-5% accuracy improvements.

Effective on challenging domain-shift scenarios.

Abstract

We introduce Transductive Infomation Maximization (TIM) for few-shot learning. Our method maximizes the mutual information between the query features and their label predictions for a given few-shot task, in conjunction with a supervision loss based on the support set. Furthermore, we propose a new alternating-direction solver for our mutual-information loss, which substantially speeds up transductive-inference convergence over gradient-based optimization, while yielding similar accuracy. TIM inference is modular: it can be used on top of any base-training feature extractor. Following standard transductive few-shot settings, our comprehensive experiments demonstrate that TIM outperforms state-of-the-art methods significantly across various datasets and networks, while used on top of a fixed feature extractor trained with simple cross-entropy on the base classes, without resorting to complex meta-learning schemes. It consistently brings between 2% and 5% improvement in accuracy over the best performing method, not only on all the well-established few-shot benchmarks but also on more challenging scenarios,with domain shifts and larger numbers of classes.