Head Gesture Recognition based on 6DOF Inertial sensor using Artificial Neural Network

This paper proposed and investigated the head gesture recognition idea based on a single inertial sensor and deep learning approach. During the experiments presented in this paper, the main aim was to evaluate and determine the best deep learning models that can accurately recognize the head movemen...

Full description

Saved in:
Bibliographic Details
Main Authors: Severin, Ionut-Cristian, Dobrea, Dan-Marius
Format: Conference Proceeding
Language:English
Subjects:
Classification algorithms
convolutional neural network
Convolutional neural networks
Deep learning
Head recognition activity (HRA)
human activity classification
Inertial sensor
LSTM
Magnetic heads
Neural Network
Testing
Time series analysis
Training
Unsupervised Machine Learning
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper proposed and investigated the head gesture recognition idea based on a single inertial sensor and deep learning approach. During the experiments presented in this paper, the main aim was to evaluate and determine the best deep learning models that can accurately recognize the head movements. The developed system can identify eight head gestures activities. The data collected in this work was done using a 6DOF inertial sensor placed on a headphone pair from 9 volunteers with ages between 20 and 40 years old. The results show that the best-proposed model's classification accuracy did reach a value equal to 93.64 % using raw data. In this experiment, each head gesture command's accuracy was in the range of 71.11 % to 100 %.
ISSN:2475-7861
DOI:10.1109/ISETC50328.2020.9301099