Wednesday, June 3, 2015

Then again, probably not

Brain's reaction to certain words could replace passwords (Binghampton University)
According to Sarah Laszlo, assistant professor of psychology and linguistics at Binghamton University and co-author of "Brainprint," brain biometrics are appealing because they are cancellable and cannot be stolen by malicious means the way a finger or retina can.
"Just 12 more globs and some wiring and you can check
your email!"

Image source:

When the alternative is the terrifying prospect of a stolen retina*, I guess you can't be too careful.

But, let's not get ahead of ourselves. Though there is little doubt that if any behavioral biometric can be used as a reliable identifier, evidence for that uniqueness could probably be found in the brain, measured, and used for ID purposes. Even so, brain prints as ubiquitous biometrics face every obstacle we discussed in our post, The challenges confronting any new biometric modality, and then some.

The linked article doesn't make any mention of the sensor to be used to collect brain prints, much less offer a vision for how a future identification scenario might work.

This is one of those subjects that is intensely interesting from a Ph.D. candidate's point of view (invention) but not so much from an engineering or business perspective (innovation). Brain prints as a biometric will face significant — I dare say insurmountable — challenges in finding their way into wide use as a commercial ID management application any time soon.

The 94% accuracy is an issue, too.

*See also:

Iris ≠ Retina

Iris (left); Retina (right)

In fairness, the penultimate paragraph in the article quotes Zhanpeng Jin, who brings a more moderate perspective to the piece.