SignAir - In-air signature biometric system making use of machine learning techniques

@Article {
title = {SignAir - In-air signature biometric system making use of machine learning techniques},
pages = {201-204},
keywds = {
Biometry, In-Air Signature, Industry 4.0, Cybersecurity, Android App, Android, Accelerometer, IoT Authentication.
}
abstract = {

Signature is an effective method to identify or verify individuals. With the latest technologies, it has arised the possibility of using an “in-air signature”, which is harder to forge than the classical handwritten signature thanks to the added dimension of depth. This project proposes a dynamic “in-air signature” biometric system that authenticates a user using a three dimensional signature captured by an Android device (making use of accelerometer and gyroscope). The in-air signature is then sent to the system in which the user will be authenticated, where it will be compared with its reconstruction, obtained by passing it through the user’s corresponding autoencoder. The result of the comparison will authenticate or reject the user. The outcome of the project is a complete biometrical system, from the capture of the signature to the authentication or identification of the user. This system achieves an equal error rate of 0.075%.

}
isbn = {978-84-88734-13-6},
date = {2022-12-17},
}