Search

• A detailed review of latest deep metric learning methods for few-shot image classification. • More than 60 methods published from 2018 to 2022 are surveyed. • Methods are divided according to three stages of deep metric learning. • Challenges and future directions of few-shot image classification are discussed. Few-shot image classification is a challenging problem that aims to achieve the human level of recognition based only on a small number of training images. One main solution to few-shot image classification is deep metric learning. These methods, by classifying unseen samples according to their distances to few seen samples in an embedding space learned by powerful deep neural networks, can avoid overfitting to few training images in few-shot image classification and have achieved the state-of-the-art performance. In this paper, we provide an up-to-date review of deep metric learning methods for few-shot image classification from 2018 to 2022 and categorize them into three groups according to three stages of metric learning, namely learning feature embeddings, learning class representations, and learning distance measures. Under this taxonomy, we identify the trends of transitioning from learning task-agnostic features to task-specific features, from simple computation of prototypes to computing task-dependent prototypes or learning prototypes, from using analytical distance or similarity measures to learning similarities through convolutional or graph neural networks. Finally, we discuss the current challenges and future directions of few-shot deep metric learning from the perspectives of effectiveness, optimization and applicability, and summarize their applications to real-world computer vision tasks. [ABSTRACT FROM AUTHOR]

• Propose two new approaches (DAMSD and DAMSDI) to HSI subpixel target detection. • Address scarcity of target spectra and improve subpixel target subspace estimation. • Provide and prove two ways of synthesising target-background mixed spectra. The performance of subspace-based methods such as matched subspace detector (MSD) and MSD with interaction effects (MSDinter) heavily depends on the background subspace and the target subspace. Nonetheless, constructing a representative target subspace is challenging due to the limited availability of target spectra in a collected hyperspectral image. In this paper, we propose two new hyperspectral target detection methods termed data-augmented MSD (DAMSD) and data-augmented MSDinter (DAMSDI) that can effectively solve the scarcity problem of target spectra and from which a representative target-background mixed subspace can be learned. We first synthesise target-background mixed spectra based on classical hyperspectral mixing models and then learn a target-background mixed subspace via principal component analysis. Compared with MSD and MSDinter, the learned mixed subspace is more representative as spectral variability of target spectra is explained to the largest extent and it leads to an improvement in computational speed and numerical stability. We demonstrate the efficacy of DAMSD and DAMSDI for subpixel target detection on two public hyperspectral image datasets. [ABSTRACT FROM AUTHOR]

Metric-based methods are one of the most common methods to solve the problem of few-shot image classification. However, traditional metric-based few-shot methods suffer from overfitting and local feature misalignment. The recently proposed feature reconstruction-based approach, which reconstructs query image features from the support set features of a given class and compares the distance between the original query features and the reconstructed query features as the classification criterion, effectively solves the feature misalignment problem. However, the issue of overfitting still has not been considered. To this end, we propose a self-reconstruction metric module for diversifying query features and a restrained cross-entropy loss for avoiding over-confident predictions. By introducing them, the proposed self-reconstruction network can effectively alleviate overfitting. Extensive experiments on five benchmark fine-grained datasets demonstrate that our proposed method achieves state-of-the-art performance on both 5-way 1-shot and 5-way 5-shot classification tasks. Code is available at https://github.com/liz-lut/SRM-main. • A novel fine-grained few-shot classification method is proposed to alleviate overfitting. • The proposed method uses self-reconstruction to increase the diversity of query features. • The proposed method introduces a restrained cross-entropy loss. • The proposed method generalizes well across five benchmark datasets. [ABSTRACT FROM AUTHOR]