Artificial Intelligence Depot
Visiting guest. Why not sign in?
News, knowledge and discussion for the AI enthusiast.
FEATURES COMMUNITY KNOWLEDGE SEARCH  
Flesh and Machines: How Robots Will Change Us
Antitechnologists may shudder at the story line, but readers interested in the astounding possibilities of the digital age will be fascinated by Brooks's vision of what is and what will be.
More information at Amazon US UK

Reply to Message

Not registered yet?

The AI Depot has a focused community of friendly users. Rather than let anyone abuse the site at the brink of promiscuity, we prefer to let only those with an active interest participate... this simply requires registering.

Why not sign up!

Joining the site's community is completely free. You can then post messages freely, and customise your personal profile at will. Specific privileges will also be granted to you, like being able to access printer-friendly articles without restrictions. So, why not register?

Username:
Password:
Subject:
Email me when someone replies.
Body:

Parent Message

Face Slap

Wired has posted an article about face recognition in airports, and recent tests that have taken place in Palm Beach; Airport Face Scanner Failed. In the article, the operating conditions and training data is briefly described. This is very interesting, as it shows the failure of the solution designed by Visionics in such a practical application. The company naturally claims that the system wasn't used right (I guess they wanted people to stand in front of the camera for 10 seconds ;) In such a high-profile test, the company should have sent experts on location to help use the system correctly!

Though this technology was hyped a lot after September, it seems to be falling flat on its face -- pardon the pun. Or at least Visionics is having trouble; lets see if Viisage can take advantage of the slip-up.

935 posts.
Thursday 16 May, 14:04
Reply
Facing facts

Face recognition in real environments is tricky and I think there have been other examples here in the UK of recognition systems being installed and then failing to work properly.

I think the commercial systems use a variety of techniques for identifying faces including eigenfaces and wavelet encoding, but the main problems are that in real situations (not ideal laboratory conditions with constant illumination) lighting can vary a great deal and there are shadows which confuse the system and so on.

Maybe in five years time these sort of systems will be effective. I can remember in the mid 1990s the UK police authorities tried to install an automatic fingerprint recognition system which also failed spectacularly, but now at my company we have a fingerprint based time and attendance system which works with a high level of reliability. The difference between a useless and practical system in the fingerprint case seems to be purely processing speed.

- Bob

136 posts.
Friday 17 May, 02:36
Reply
Faces, Gaits & Behaviours

Indeed, it's tricky business, but that didn't stop them from riding the popularity wave and promoting their system... now they're paying the consequences.

I have to confess, I don't like the fact that security systems are so single-track. They'd benefit much more from combining gait recognition, behaviour monitoring (with those smart cameras) and face detection all together. It seems a much more reliable approach to me...

935 posts.
Sunday 19 May, 13:46
Reply
Children as Inspiration For Rodney's Image Recognition?

Science Daily has this interesting article about face recognition in children. It describes research showing that kids learn to recognise and differentiate faces at an early age.

Study Suggests Infants "Tune In" To Familiar Face Groups

The article is biologically based, so the technical details are sparse. But it seems to me like there might be some good ideas there.

The infants learn the features that distinguish faces. So, for your decision tree learning algorithm, you could have a first pass that finds the set of features that maximises the entropy (difference between samples). This could be done with Genetic Algorithms (selecting expert features) or maybe even Genetic Programming (automatically creating compound features). Normal learning could then be done after that...

What do you reckon? Am I onto a winner? I've not done much face recognition, so I'm not sure...

935 posts.
Monday 20 May, 16:21
Reply
Image recognition

This article supports preexisting research in that it sugests that face recognition is a flexible ability based on frequency of observation. This is entirely in line with studies done in the 1980s on the somatosensory cortex of monkeys, where the area of cortex responsible for processing tactile inputs from various parts of the body is proportional to the usage of that part of the body. The interesting part of this research was that it was shown that the somatosensory cortex does not remain static but is instead a dynamic competitive system. For example, if you use one finger more than any other the area of your brain dedicated to picking up signals from that finger becomes relatively larger, and conversely if that finger were to unfortunately be chopped off then the area of brain dedicated to it shrinks as surrounding areas colonise the unused space.

This article would suggest that what's happening in the visual domain is probably no different from what happens with tactile senses. You can model this sort of system using kohonen networks, or alternatively as more elaborate schemes such as the one Gerald Edelman used in his book "Neural Darwinism". The central principle is that these are competitive systems based on a frequency of use or frequency of observation basis.

You've made a good point in that there is as much value in searching for features which are *not* important as those that are. This is something which I havn't really taken into account in the systems which I've devised so far. Most systems which attempt to find patterns look for commonalities, not differences.

Just reading the Kurzweil/Wolfram review I tried to apply a cellular automata style technique to the problem of visual recognition within the ORAC system. Here there are a number of simple computational elements running in parallel. Like Wolfram my thinking was that the actual complexity of the computational elements is largely unrelated to the complexity of the end result. You could use simple logic gates, or devise very complicated simulated neurons.

- Bob

136 posts.
Monday 20 May, 17:03
Reply

Back to the Artificial Intelligence Depot.