GAN Based Data Augmentation for Indoor Localization Using Labeled and Unlabeled Data

Wafa Njima, Marwa Chafii, Raed M. Shubair

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Machine learning techniques allow accurate indoor localization with low online complexity. However, a large amount of collected data samples is needed to properly train a deep neural network (DNN) model used for localization. In this paper, we propose to generate fake fingerprints using generative adversarial networks (GANs) based on a small amount of collected data samples. We consider an indoor scenario where collected labeled data samples are rare and insufficient to generate fake samples of a good multitude and diversity in order to provide a good localization accuracy. Thus, both labeled and unlabeled fingerprints are provided to the GAN so that more realistic fake data samples are generated. Then, a DNN model is trained on mixed dataset comprising real collected labeled and pseudo-labeled fingerprints as well as fake generated pseudo-labeled fingerprints. The data augmentation based on real measurements leads to a mean localization accuracy improvement of 9.66% in comparison to the conventional semi-supervised localization algorithm.

Original languageEnglish (US)
Title of host publication2021 International Balkan Conference on Communications and Networking, BalkanCom 2021
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages36-39
Number of pages4
ISBN (Electronic)9781665402583
DOIs
StatePublished - Sep 20 2021
Event4th International Balkan Conference on Communications and Networking, BalkanCom 2021 - Novi Sad, Serbia
Duration: Sep 20 2021Sep 22 2021

Publication series

Name2021 International Balkan Conference on Communications and Networking, BalkanCom 2021

Conference

Conference4th International Balkan Conference on Communications and Networking, BalkanCom 2021
Country/TerritorySerbia
CityNovi Sad
Period9/20/219/22/21

Keywords

  • deep neural network (DNN)
  • generative adversarial network (GAN)
  • Indoor localization
  • received signal strength indicator (RSSI)
  • semi-supervised learning

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Networks and Communications
  • Hardware and Architecture
  • Information Systems and Management
  • Safety, Risk, Reliability and Quality

Fingerprint

Dive into the research topics of 'GAN Based Data Augmentation for Indoor Localization Using Labeled and Unlabeled Data'. Together they form a unique fingerprint.

Cite this