01
Identity evidence
Compare a captured face with an approved reference using one-to-one verification.
TED UNIVERSITY / CMPE 492
Senior design project 2026–2027
Trust begins
with living signals.
Exploring biometric verification through the fusion of facial identity, visual attack detection and physiological evidence.
BioAuth is a senior research project investigating how signal quality and attack awareness can support more reliable verification decisions.
Quality-aware. Attack-aware. Explainable by design.
BIOMETRIC EVIDENCEFIG. 01 / CONCEPT
rPPG signalILLUSTRATIVE
Face verification Visual PAD Physiology
THE PROJECTBiometric verification & security
CURRENT MILESTONE Low-level design · v1 available
Follow our progress 01THE RESEARCH QUESTION
A familiar face is only part of the story.
Face matching asks who is in an image. It does not establish whether a capture comes from a person, a replay or synthetic media.
BioAuth investigates a quality- and attack-aware fusion engine that considers identity, presentation and physiology together—and requests another capture when the evidence is insufficient.
Understand the approach02
Presentation evidence
Evaluate visual attack indicators independently from whether the face matches.
03
Physiological evidence
Assess pulse-related colour variations and their reliability with CHROM and POS.
Research hypothesis: improve the trade-off between attack acceptance, genuine-user rejection, retries and latency. Results remain to be validated.
02PROJECT DOCUMENTATION
The work, documented.
Design decisions, technical detail and course deliverables in one place.
CMPE 492 / Version 1.0
16 October 2026CMPE 492 / Testing
Test Plan Report
Due 20 November 2026CMPE 492 / Final delivery
Final Report
Due 31 December 2026 Public course deliverables · CMPE 492TODO & backlog archive
03 THE SEMESTER AHEAD
From a detailed design
to a tested system.
Follow the project through design refinement, implementation, controlled evaluation and the final demonstration.
View milestones & backlog16 OCTPublished
Low-Level Design
Version 1 available
20 NOV
Test Plan
Evaluation methodology & test cases
31 DEC
Final Report
Implementation & measured results
JAN '27
Presentation & Demo
Exact date to be announced
04THE PEOPLE BEHIND THE PROJECT
Four students. A shared research goal.
EK+
Esra Eda Kılıç
Computer Engineering
TED UniversityET+
Enes Taşçı
Computer Engineering
TED UniversityÖA+
Ömer Taylan Alemdar
Computer Engineering
TED UniversityBK+
Batuhan Mert Korkmaz
Computer Engineering
TED UniversitySUPERVISED BYDr. Ulaş GüleçDepartment of Computer Engineering · TED University
Curious about the implementation?Explore the source code and follow the project as it develops.