The search functionality is under construction.
The search functionality is under construction.

Open Access
Personalized Food Image Classifier Considering Time-Dependent and Item-Dependent Food Distribution

Qing YU, Masashi ANZAWA, Sosuke AMANO, Kiyoharu AIZAWA

  • Full Text Views

    89

  • Cite this
  • Free PDF (2MB)

Summary :

Since the development of food diaries could enable people to develop healthy eating habits, food image recognition is in high demand to reduce the effort in food recording. Previous studies have worked on this challenging domain with datasets having fixed numbers of samples and classes. However, in the real-world setting, it is impossible to include all of the foods in the database because the number of classes of foods is large and increases continually. In addition to that, inter-class similarity and intra-class diversity also bring difficulties to the recognition. In this paper, we solve these problems by using deep convolutional neural network features to build a personalized classifier which incrementally learns the user's data and adapts to the user's eating habit. As a result, we achieved the state-of-the-art accuracy of food image recognition by the personalization of 300 food records per user.

Publication
IEICE TRANSACTIONS on Information Vol.E102-D No.11 pp.2120-2126
Publication Date
2019/11/01
Publicized
2019/06/21
Online ISSN
1745-1361
DOI
10.1587/transinf.2019PCP0005
Type of Manuscript
Special Section PAPER (Special Section on Picture Coding and Image Media Processing)
Category

Authors

Qing YU
  The University of Tokyo
Masashi ANZAWA
  The University of Tokyo
Sosuke AMANO
  The University of Tokyo,foo.log Inc.
Kiyoharu AIZAWA
  The University of Tokyo

Keyword