Accelerating Model Validation by Reconstructing Image Sets Using GAN

In recent years, the field of image recognition has been advancing rapidly, and various methods have been developed. However, it is necessary to perform many validations because the appropriate network structure differs depending on the type of image. In many examples, models have been validated usi...

Full description

Saved in:
Bibliographic Details
Published in2021 International Conference on Computational Science and Computational Intelligence (CSCI) pp. 1736 - 1739
Main Authors Hatano, Toshiki, Kiriyama, Taro, Tsuneda, Toi, Yamane, Satoshi
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.12.2021
Subjects
Color
Computational intelligence
Computational modeling
DNN
GAN
Generative adversarial networks
Image recognition
Image reconstruction
Scientific computing
Online AccessGet full text

Cover

More Information
Summary:In recent years, the field of image recognition has been advancing rapidly, and various methods have been developed. However, it is necessary to perform many validations because the appropriate network structure differs depending on the type of image. In many examples, models have been validated using all the original image dataset, which requires a lot of effort. In this study, we used CGAN to reconstruct a small image dataset and succeeded in significantly reducing the model validation time.
DOI:10.1109/CSCI54926.2021.00329