Cited 0 time in webofscience Cited 7 time in scopus

Generative Adversarial Networks to Augment Micro-Doppler Signatures for the Classification of Human Activity

Title
Generative Adversarial Networks to Augment Micro-Doppler Signatures for the Classification of Human Activity
Authors
Alnujaim, IbrahimOh, DaegunKim, Youngwook
DGIST Authors
Oh, Daegun
Issue Date
2019-08-02
Citation
39th IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2019, 9459-9461
Type
Conference
ISBN
9781538691540
Abstract
Collecting a large amount of data for radar requires a significant amount of time, labor, and money. In deep convolutional neural networks, a small dataset causes the problem of overfitting. We herein introduce the employment of data augmentation using generative adversarial networks (GANs) to solve the data deficiency problem. In this study, we tested the feasibility of using generative adversarial networks to generate micro-Doppler signatures for seven human activities. Moreover, we use produced fake images to train deep convolutional neural networks. We found that the use of augmented data improves classification accuracy. In addition, the quality of GAN output was evaluated in terms of classification accuracy. © 2019 IEEE.
URI
http://hdl.handle.net/20.500.11750/11441
DOI
10.1109/IGARSS.2019.8898073
Publisher
Institute of Electrical and Electronics Engineers Inc.
Related Researcher
Files:
There are no files associated with this item.
Collection:
Division of Intelligent Robot2. Conference Papers


qrcode mendeley

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE