Publication
Title
Bio-inspired gesture recognition with baffled transducers using temporal and spectral features
Author
Abstract
Echolocating bats can provide engineers with tremendous inspiration for constructing active ultrasonic sensors for airborne applications. Previous research has demonstrated that the recognition of hand gestures can be facilitated by means of ultrasonic sensing, often relying on classical engineering principles. In this paper we merge the insights from research into biological echolocation systems with gesture recognition, and present a gesture recognition sensor inspired by the echolocation system of bats, using spatiospectral features induced by the geometric shape of baffled microphones. We use the spatiospectral features in combination with a support vector machine classifier as gesture classification mechanism. We show the efficacy of the proposed approach using experimental data gathered from twenty persons performing four different gestures.
Language
English
Source (book)
2023 IEEE SENSORS, 29 October - 01 November, 2023, Vienna, Austria
Publication
IEEE , 2023
ISBN
979-83-503-0387-2
DOI
10.1109/SENSORS56945.2023.10324935
Volume/pages
p. 1-4
Full text (Publisher's DOI)
UAntwerpen
Faculty/Department
Research group
Publication type
Subject
Affiliation
Publications with a UAntwerp address
External links
Record
Identifier
Creation 19.01.2024
Last edited 20.01.2024
To cite this reference