Confusable Learning for Large-class Few-Shot Classification

TL;DR

This paper introduces Confusable Learning, a novel approach for large-class few-shot image classification that emphasizes confusable classes using a dynamic confusion matrix, improving performance over existing methods.

Contribution

It proposes a biased learning paradigm that dynamically identifies and emphasizes confusable classes, enhancing meta-learning algorithms for large-class few-shot tasks.

Findings

Outperforms state-of-the-art baselines on Omniglot, Fungi, and ImageNet.

Effectively identifies confusable classes to improve classification accuracy.

Demonstrates robustness in large-class few-shot scenarios.

Abstract

Few-shot image classification is challenging due to the lack of ample samples in each class. Such a challenge becomes even tougher when the number of classes is very large, i.e., the large-class few-shot scenario. In this novel scenario, existing approaches do not perform well because they ignore confusable classes, namely similar classes that are difficult to distinguish from each other. These classes carry more information. In this paper, we propose a biased learning paradigm called Confusable Learning, which focuses more on confusable classes. Our method can be applied to mainstream meta-learning algorithms. Specifically, our method maintains a dynamically updating confusion matrix, which analyzes confusable classes in the dataset. Such a confusion matrix helps meta learners to emphasize on confusable classes. Comprehensive experiments on Omniglot, Fungi, and ImageNet demonstrate the efficacy of our method over state-of-the-art baselines.