中国邮电高校学报(英文) ›› 2022, Vol. 29 ›› Issue (3): 15-24.doi: 10.19682/j.cnki.1005-8885.2022.1015

• • 上一篇    下一篇

Ensemble relation network with multi-level measure

Li Xiaoxu, Qu Xue, Cao Jie
  

  1. 1. College of Computer and Communication, Lanzhou University of Technology, Lanzhou 730050, China
    2. College of Electrical and Information Engineering, Lanzhou University of Technology, Lanzhou 730050, China
    3. Engineering Research Center of Urban Railway Transportation of Gansu Province, Lanzhou 730050, China

  • 收稿日期:2022-04-16 修回日期:2022-05-26 出版日期:2022-06-30 发布日期:2022-06-30
  • 通讯作者: 曹洁 E-mail:caoj@lut.edu.cn
  • 基金资助:
    This work was supported by the National Natural Science Foundation of China (62176110, 62111530146, 61906080), and Young Doctoral Fund of Education Department of Gansu
    Province (2021QB-038)

Ensemble relation network with multi-level measure

Li Xiaoxu, Qu Xue, Cao Jie   

  1. 1. College of Computer and Communication, Lanzhou University of Technology, Lanzhou 730050, China
    2. College of Electrical and Information Engineering, Lanzhou University of Technology, Lanzhou 730050, China
    3. Engineering Research Center of Urban Railway Transportation of Gansu Province, Lanzhou 730050, China

  • Received:2022-04-16 Revised:2022-05-26 Online:2022-06-30 Published:2022-06-30
  • Supported by:
    This work was supported by the National Natural Science Foundation of China (62176110, 62111530146, 61906080), and Young Doctoral Fund of Education Department of Gansu
    Province (2021QB-038)

摘要: Fine-grained few-shot learning is a difficult task in image classification. The reason is that the discriminative
features of fine-grained images are often located in local areas of the image, while most of the existing few-shot learning image classification methods only use top-level features and adopt a single measure. In that way, the local features of the sample cannot be learned well. In response to this problem, ensemble relation network with multi-level measure (ERN-MM) is proposed in this paper. It adds the relation modules in the shallow feature space to compare the similarity between the samples in the local features, and finally integrates the similarity scores from the feature spaces to assign the label of the query samples. So the proposed method ERN-MM can use local details and global information of different grains. Experimental results on different fine-grained datasets show that the proposed method achieves good classification performance and also proves its rationality.

关键词: fine-grained image classification, few-shot learning, local feature learning, metric learning

Abstract: Fine-grained few-shot learning is a difficult task in image classification. The reason is that the discriminative
features of fine-grained images are often located in local areas of the image, while most of the existing few-shot learning image classification methods only use top-level features and adopt a single measure. In that way, the local features of the sample cannot be learned well. In response to this problem, ensemble relation network with multi-level measure (ERN-MM) is proposed in this paper. It adds the relation modules in the shallow feature space to compare the similarity between the samples in the local features, and finally integrates the similarity scores from the feature spaces to assign the label of the query samples. So the proposed method ERN-MM can use local details and global information of different grains. Experimental results on different fine-grained datasets show that the proposed method achieves good classification performance and also proves its rationality.

Key words: fine-grained image classification, few-shot learning, local feature learning, metric learning