<b>A Model Ensemble Approach for Few-Shot Learning Using Aggregated Classifiers</b>

• KIKUCHI Toshiki

  Keio University

• OZASA Yuko

  Keio University

Bibliographic Information

Other Title

• A Model Ensemble Approach for Few-Shot Learning Using Aggregated Classifiers

Abstract

Despite the recent success in deep neural networks on the visual domain, we need a large amount of data to train the networks. Previous works addressed this issue as the few-shot learning which is the task to identify the class of an example in new classes not seen in a training phase with only a few examples of each new class. Some methods performed well on the few-shot tasks, but need a complex architecture and/or specialized loss functions, such as metric loss, meta learner, and memory. In this paper,we evaluate the performance of the ensemble approach aggregating a huge number of simple neural network models (up to 128 models) on standard few-shot datasets. Surprisingly, although the approach is simple, our experimental results show that the ensemble approach is competitive with state-of-the-art methods among similar architecture methods in some settings.

Journal

• IIEEJ Transactions on Image Electronics and Visual Computing

  IIEEJ Transactions on Image Electronics and Visual Computing 7 (2), 97-105, 2019-12-15

  The Institute of Image Electronics Engineers of Japan

Keywords

• few-shot classification
• model ensemble
• deep learning

Details

• CRID: 1390850490583961984
• NII Article ID: 130008012590
• NII Book ID: AA12661628
• DOI: 10.11371/tievciieej.7.2_97
• ISSN: 21881901
• ISSN: 2188191X
• ISSN: 21881898
• NDL BIB ID: 030259934
• Web Site: https://ndlsearch.ndl.go.jp/books/R000000004-I030259934
• Text Lang: en
• Data Source:
  • JaLC
  • NDL
  • CiNii Articles
  • KAKEN
• Abstract License Flag: Disallowed