Multi-level Metric Learning for Few-shot Image Recognition

TL;DR

This paper introduces a multi-level metric learning approach that combines pixel, part, and global features for improved few-shot image recognition, demonstrating superior performance over existing methods.

Contribution

It proposes a novel multi-level metric learning framework that captures diverse feature similarities and a part-level embedding adaptation with graph to enhance few-shot learning.

Findings

Outperforms state-of-the-art methods on popular datasets.

Effectively captures multi-level feature similarities.

Improves classification accuracy in few-shot scenarios.

Abstract

Few-shot learning is devoted to training a model on few samples. Most of these approaches learn a model based on a pixel-level or global-level feature representation. However, using global features may lose local information, and using pixel-level features may lose the contextual semantics of the image. Moreover, such works can only measure the relations between them on a single level, which is not comprehensive and effective. And if query images can simultaneously be well classified via three distinct level similarity metrics, the query images within a class can be more tightly distributed in a smaller feature space, generating more discriminative feature maps. Motivated by this, we propose a novel Part-level Embedding Adaptation with Graph (PEAG) method to generate task-specific features. Moreover, a Multi-level Metric Learning (MML) method is proposed, which not only calculates the pixel-level similarity but also considers the similarity of part-level features and global-level features. Extensive experiments on popular few-shot image recognition datasets prove the effectiveness of our method compared with the state-of-the-art methods. Our code is available at \url{https://github.com/chenhaoxing/M2L}.