...

Gesture Dynamics with skin conductivity

by user

on
10

views

Report

Comments

Transcript

Gesture Dynamics with skin conductivity
GESTURE DYNAMICS WITH SKIN CONDUCTIVITY - A MULTIMODAL APPROACH TO
BIOMETRIC AUTHENTICATION
VÍTOR JÚLIO DA SILVA E SÁ *
Supervisor: Henrique Dinis dos Santos
University of Minho
School of Engineering
Algoritmi Centre
* [email protected]
Keywords
Problem
Research Plan
Security, authentication, biometry, gesture dynamics, cognitive
biometrics, skin conductivity, multimodal.
In the current solution we have identified the following problems:
replicability is possible even with the common biometric technology;
requirement of continuous authentication; constant need to improve
the accuracy of the systems (information systems with more users
imply a need to lower error rates); constant need (inherent in the
existence of a business) to lower the cost; need to keep the mobility of
the systems, with restrictions on the type of hardware needed
(dimension, energy, etc.) and on the use conditions (light, noise, etc.).
For better exposure of the tasks involved, results, goals, resources
and timing, the research plan was developed in a project management
tool, with the corresponding set of deliverables, activities, effort,
duration, dependencies and milestones.
Motivation
There different ways to recognize an individual.
There is recent work in cognitive biometrics, particularly in the
conductivity of the skin that seems to indicate that this technology has
potential (with or without activation of the knowledge component). The
current size of the conductivity sensors enables its integration into
mobile devices. There is also work done and consolidated in gesture
dynamics.
The latter methods are known as biometrics, which overcomes the
problem of transmissibility that the first two ways have.
Thus, we intend to prove the following: it is possible to implement, with
comparative advantages, a multimodal biometric authentication
solution using gesture dynamics with skin conductivity.
Biometrics classification:
Methodological Approach
Given the nature of this work the research methodology is based on a
proof of concept with a study of acceptance. For that, it will be
developed a prototype for graphical data collection and a prototype for
dermoelectric data collection, for further definition of algorithms that
determine authentication patterns.
However, biometrics is not foolproof because it can be replicated.
Today there are several fraudulent ways to replicate a biometric
characteristic, and this raises new challenges.
A potential solution lies in cognitive biometrics, which uses biological
signals representative of the mental and emotional states for the
authentication of users.
The main challenges will be to find a good algorithm for each of the
biometrics and find the thresholds that combine successfully these
same algorithms. This does not correspond to the simple junction of
the two since it will take advantage of the synergistic multimodal
integration or even, in some situations, the alternate one.
Later we will proceed to the test of the defined algorithms, with half of
the data to test and half of them to attack.
Developed the solution, the aim is also to evaluate the user
acceptance of such a system. We plan to use the Technology
Acceptance Model (TAM) for its wide dissemination and acceptance in
this context.
Publications
Borges, D.; Sá, V. J.; Magalhães, S. T.; Santos, H. D. 2011. “Study of
the perception on the biometric technology by the Portuguese
citizens”. Proceedings of the 7th ICGS3 / 4th e-Democracy Joint
Conferences, Salónica, Grécia.
Cerqueira, I.; Sá, V. J.; Magalhães, S. T. 2011. “Study of The
Perception on The Portuguese Citizen Card and Electronic Signature”.
Proceedings of the 7th ICGS3 / 4th e-Democracy Joint Conferences,
Salónica, Grécia.
Faria, L.; Sá, V. J.; Magalhães, S. T. 2011. “Biometrias Multimodais
Cognitivas - Uma nova abordagem no reconhecimento de
utilizadores”. In Actas da 6ª Conferência Ibérica de Sistemas e
Tecnologias de Informação, Chaves, Portugal.
Hildebrand, A. and Sá, V. J. 2001. “EMBASSI: Electronic Multimedia
and Service Assistance”. In: Heuer, A., Kirste, T. (eds.) Intelligent
Interactive Assistance and Mobile Multimedia Computing, ISBN 3935319-75-4, pp. 50-59, Neuer Hochschulschriftenverlag, Rostock.
Sá, V. J.; Malerczyk, C.; Schnaider, M. (2002). Vision Based
Interaction Within a Multimodal Framework. Selected Readings in
Computer Graphics 2001, ISBN: 3-8167-6163-1, Fraunhofer IRB
Verlag, Stuttgart.
Uma Escola a Reinventar o Futuro – Semana da Escola de Engenharia - 20 a 26 de Outubro de 2011
Fly UP