Deep metric learning for few-shot image classification: A Review of recent developments

被引:94
作者
Li, Xiaoxu [1 ,2 ]
Yang, Xiaochen [3 ]
Ma, Zhanyu [2 ]
Xue, Jing-Hao [4 ]
机构
[1] Lanzhou Univ Technol, Sch Comp & Commun, Lanzhou 730050, Peoples R China
[2] Beijing Univ Posts & Telecommun, Sch Artif Intelligence, Pattern Recognit & Intelligent Syst Lab, Beijing 100876, Peoples R China
[3] Univ Glasgow, Sch Math & Stat, Glasgow G12 8QQ, Scotland
[4] UCL, Dept Stat Sci, London WC1E 6BT, England
基金
北京市自然科学基金;
关键词
Few -shot learning; Metric learning; Image classification; Deep neural networks; few-shot; one-shot learning; one-shot classification; one-shot image recognition; NETWORK;
D O I
10.1016/j.patcog.2023.109381
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
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.(c) 2023 Elsevier Ltd. All rights reserved.
引用
收藏
页数:15
相关论文
共 50 条
  • [41] Automated classification of polyps using deep learning architectures and few-shot learning
    Krenzer, Adrian
    Heil, Stefan
    Fitting, Daniel
    Matti, Safa
    Zoller, Wolfram G.
    Hann, Alexander
    Puppe, Frank
    BMC MEDICAL IMAGING, 2023, 23 (01)
  • [42] Automated classification of polyps using deep learning architectures and few-shot learning
    Adrian Krenzer
    Stefan Heil
    Daniel Fitting
    Safa Matti
    Wolfram G. Zoller
    Alexander Hann
    Frank Puppe
    BMC Medical Imaging, 23
  • [43] Classification of Marine Plankton Based on Few-shot Learning
    Jin Guo
    Jihong Guan
    Arabian Journal for Science and Engineering, 2021, 46 : 9253 - 9262
  • [44] Classification of Marine Plankton Based on Few-shot Learning
    Guo, Jin
    Guan, Jihong
    ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING, 2021, 46 (09) : 9253 - 9262
  • [45] Multiscale attention for few-shot image classification
    Zhou, Tong
    Dong, Changyin
    Song, Junshu
    Zhang, Zhiqiang
    Wang, Zhen
    Chang, Bo
    Chen, Dechun
    COMPUTATIONAL INTELLIGENCE, 2024, 40 (02)
  • [46] ICCL: Independent and Correlative Correspondence Learning for few-shot image classification
    Zheng, Zijun
    Wu, Heng
    Lv, Laishui
    Ye, Hailiang
    Zhang, Changchun
    Yu, Gaohang
    KNOWLEDGE-BASED SYSTEMS, 2023, 266
  • [47] Generalized few-shot learning for crop hyperspectral image precise classification
    Yuan, Hao-tian
    Huang, Ke-kun
    Duan, Jie-li
    Lai, Li-qian
    Yu, Jia-xiang
    Huang, Chao-wei
    Yang, Zhou
    COMPUTERS AND ELECTRONICS IN AGRICULTURE, 2024, 227
  • [48] Rare Data Image Classification System Using Few-Shot Learning
    Lee, Juhyeok
    Kim, Mihui
    ELECTRONICS, 2024, 13 (19)
  • [49] Contrastive prototype learning with semantic patchmix for few-shot image classification
    Dong, Mengping
    Lei, Fei
    Li, Zhenbo
    Liu, Xue
    ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE, 2025, 142
  • [50] Multiview Calibrated Prototype Learning for Few-Shot Hyperspectral Image Classification
    Yu, Chunyan
    Gong, Baoyu
    Song, Meiping
    Zhao, Enyu
    Chang, Chein-, I
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2022, 60