Re: Biometrics at infragard *the only thing missing is u!*

[Job de Haas <joblists () itsx com>]

Hi,

Thanks for the overview, interesting to see how busy people are with it.

> One thing I noticed is a lack of "adversarial testing". This is 

This is a key point. There seems to be a complete lack of it actually.
And there has been for years (> 15). For some reasons all systems are
qualified by there percentage of properly identifying persons based on
their features. Never you see how well a system was able to identify a
fake between real ones.

> where people are free to paint their face. Even the highly vaunted 
> fingerprint recognition teams are still reeling from the "gummy bear" 
> attack where a Japanese scientist bypassed a fingerprint biometric with 
> a piece of 10 dollar gelatin.

A friend of mine has been doing this for a long time:
http://www.keuning.com/biometry/

I think he is able to make a fake from a real print in about 15 minutes.
To get more reliable fake prints you need more time, but he said he
never saw a fingerprint system he could not fool. Despite 'resistance
measurement' 'pulse detection' or what ever.

My gut feeling it will be at least the same for iris detection. I've
heard some simple tests with color printed irises, but someone in the
contactlense business may have far better equipment to fool those
systems.

Job

_______________________________________________
Dailydave mailing list
Dailydave () lists immunitysec com
https://lists.immunitysec.com/mailman/listinfo/dailydave