Biometrics: Identity Verification in a Networked World Transparent Multimodal Biometric System for HD Multimedia Conference Name: Laith Abbadi, Abbas Javadtalab
Feb 24, 2016
Biometrics: Identity Verification in a Networked World
Transparent Multimodal Biometric System for HD Multimedia Conference
Name: Laith Abbadi, Abbas Javadtalab
Introduction
Identification in electronic networks (e-Identification) is a main topic in e-world (e-commerce and e-government) [1].
A biometric system recognizes a person based on physiological or behavioral characteristics of the person.
A multimodal biometric system is a system that combines two or more of the biometric characteristics.
Purpose
The purpose of the paper is to verify the identity of a person during a video conference using a transparent multimodal biometrics.
The paper will propose a multimodal biometric system using face and ear recognition and without the user’s interaction.
Security Proposed Framework For practical use, the framework is aiming to
satisfy a set of criteria include: (i)Ease of deployment:
▪ A system can be deployed with few additional requirements on current infrastructure and communication protocols;
(ii) Ease of use: ▪ Users have no difficulties using the system;
(iii) Security: ▪ A system should address the real security concerns in
verifying a person in the system. The security may not be perfect, but should be good enough to be user-friendly and business-driven
Authentication Type
Classification Description Example Identifier Group
Type 1 Something you know Password, PIN Knowledge based
Type 2 Something you have Token, OTP Electronic tokens
Type 3 Something you are Fingerprint, iris, face, ear Biometric (Physiological)
Type 4 Something you do Voice, Signature Biometric (Behavioral)
Table 1: Classification of authentication type
Identifier Groups Knowledge Based Identifiers
Date of Birth SIN Number Names Address PIN Numbers (passwords)
Electronic Tokens Digital tokens are the physical devices that contain digital
information for verification purposes. ▪ a. Smart Cards▪ b. One time Password Token
▪ OTP Time based Token▪ OTP Event based Token
▪ c. Radio frequency identifiers (RFID).
Identifier Groups
Biometric Identifiers Physiological Biometric Identifiers:
▪ Fingerprint ▪ Retina image▪ Iris▪ Face Recognition▪ Ear Recognition
Behavioral Biometric Identifiers: ▪ Voice▪ Dynamic signature▪ Keystrokes dynamics
Note: Using behavioral identifiers alone is not a solid solution, but they work well if they used with other types of identifiers.
Biometric System
Features Biometric for Verification
Biometric for Identification
Fingerprint √ √
Iris √
Face Recognition
√ √
Ear Recognition
√
Voice √
Dynamic Signature √
Keystrokes dynamics √
Table 2: Biometric Features
E-identifiers
Table 3: Comparison of e-identifier groups
Source: [1]
Biometric identifiers
Ear
Table 4: Biometric Identifiers [2]
Note: For Ear, research was not finished by Nov 3, 2010
HD Multimedia Conference High quality video Conferencing Used for communicating between CEO’s and
VIP’s Video Quality:1920x1080 x264:
Open source implementation of H.264 standard x264 offers faster encoding
Sample Results (CBR)
Original 1200 Kbit/s
100 Kbit/s800 Kbit/s
Multimodal Biometric Recognition system
Face Templates Face
Matching
Face Extraction
Ear Templates
Fusion
Ear Matching
Ear Extraction
Decision
Conference solution
Conference solution
Future Work
1- Apply face detection Algorithm 2- Apply ear recognition Algorithm 3- Apply voice detection
a. Oral style (such as spelling ‘aahhh’) 4- Mouth movement 5- Face Expression & Emotion (ex.
laughing) 8- Ration (face with upper body) 9- Add-ons
a. Hand shake (haptics)
Conclusion
Use of Multimodal Biometric system is more advantageous than using a mono-modal biometric system
Biometric features are unique to each person
It is feasible to have a transparent verification system using face and ear recognition
Thank You
References [1] Biometric Technology Today (BTT), June (2008) Biometrics in the retail sector page 9-
11 [2] Black Cathryn, (2008) Biometric Technology Today • January 2008page 5 [3] Clarke Roger (2007) Introduction to Information Security, February 2007 [4] Clarke, Roger (2008), ``EDI is but one element of electronic commerce'', Roger
Clarke's EC Foundation Paper http://www.anu.edu.au/people/Roger.Clarke/EC/Bled08.ht [5] Clarke, Roger (2005) Dataveillance by Governments The Technique of Computer
Matching Information Technology & People, Vol. 7 No. 2, 2005, pp. 46-85 [6] Clarke, Roger (2009) Human Identification in Information Systems: Management
Challenges and Public Policy Issues [7] Cranor L, Cytron R. Sensus:a security-conscious electronic polling system for the
Internet. Proceedings of the Hawaii International Conference on System Sciences; 2009. p. 561e70.
[8] Desmarais Norman, (2009) Body language, security and e-commerce Volume 18 . Number 1 . 2009 . pp. 61-74
[9] Granova Anna & Eloff JHP, (2004) Online banking and identity theft: who carries the risk? Computer Fraud and Security page 7-8
[10] Marshalla Angus M., Tompsett Brian, (2005) Identity theft in an online world Computer Law & Security Report (2008) 21, 128e137
[11] Monrose Fabian, Rubin Aviel D. (2009) Keystroke dynamics as a biometric for authentication. Future Generation Computer Systems 16 (2009) 351-359