Improvement for facial gestures classification to control a drone

Elena Acevedo, Antonio Acevedo, Alexa Chavez

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Over the years, unmanned aerial vehicles have had significant development and impact on different activities of the human being, such as mining, agriculture, health, military, entertainment, among others. Therefore, researchers have sought to merge this technology with others that allow it to improve its performance, make it accessible, and solve specific problems that currently cannot. One of these technologies is brain-computer interfaces, which can manipulate devices through brain signals. In this work, an Artificial Neural Network is proposed together with data preprocessing to improve the recognition of the gestures used to control an unmanned aerial vehicle. Two test algorithms, 10-fold Cross Validation and Hold Out, were applied and the efficiency results obtained were 94.14% and 98.95%, respectively.

Original languageEnglish
Title of host publication15th International Multi-Conference on Society, Cybernetics and Informatics, IMSCI 2021
EditorsNagib C. Callaos, Jeremy Horne, Belkis Sanchez, Michael Savoie
PublisherInternational Institute of Informatics and Systemics, IIIS
Pages1-5
Number of pages5
ISBN (Electronic)9781713835189
StatePublished - 2021
Event15th International Multi-Conference on Society, Cybernetics and Informatics, IMSCI 2021 - Virtual, Online
Duration: 18 Jul 202121 Jul 2021

Publication series

Name15th International Multi-Conference on Society, Cybernetics and Informatics, IMSCI 2021

Conference

Conference15th International Multi-Conference on Society, Cybernetics and Informatics, IMSCI 2021
CityVirtual, Online
Period18/07/2121/07/21

Keywords

  • Artificial intelligence
  • Classification
  • Machine learning
  • Neural networks
  • Pre-Processing

Fingerprint

Dive into the research topics of 'Improvement for facial gestures classification to control a drone'. Together they form a unique fingerprint.

Cite this