Week 12: Biometrics and authentication Flashcards
What biometric topic would your fingerprint be?
anatomy or physiology
What biometric topic would your handwriting be?
Skill or behaviour
What is voice or gait?
A mixture of both anatomy and physiology and Skill or behaviour
What are the four types of biometrics a fingerprint can be?
Island
dot
bifurcation
ridge ending
What is a fraud rate?
False acceptance
What is insult rate?
False Rejects
What is an example of feature extraction?
Sensor technology
What some examples of future technology?
Authentication Aura: User authentication within the
Internet of Things
Device centric multi-sensor gait recognition
Smart watch physical activity recognition
Facial recognition can be enhanced
with special hardware, how?
In 2017, Apple
introduced it on the iPhone X, in which
a dot projector paints your face with
tens of thousands of dots and a
camera reads them. This deals with
makeup, some sunglasses and facial
hair, and was claimed to have a false
acceptance rate of one in a million –
as opposed to one in 50,000 for the
fingerprint reader that previous
iPhones used.
Why aren’t really good extraction methods used?
Cost of sensor
Maturity of the hardware varies considerably (e.g. fingerprint sensors and leftover fingerprint flaw)
What is a liveness test? How can it be bypassed
android facial detection required end user to blink
gif images / cut out eyes from a picture
What is the curse of dimensionality?
More features, more techniques = larger classifier, increased processing time
Why are specific traits extracted?
Extracted characteristics need to be the same for consistency and accuracy
Behavioural biometric template changes more often than physiological
Why are specific traits extracted?
Extracted characteristics need to be the same for consistency and accuracy
Behavioural biometric template changes more often than physiological
To increase security would you wnt a higher rejection rate or acceptance rate?
Rejection rate
