A camera that uses AI to work out what gender the user is could have all sorts of applications in marketing, market research or just tailoring an interaction to take account of gender.

A Spanish team from the Computational Intelligence Group based at the Universidad Politécnica de Madrid have has developed a system that analyses a video signal in real time and calculates the gender of the faces pictured in the images - and they have obtained a Spanish patent on the technique.

The device is a video camera that allows a face detector to capture a 25x25 pixel cropped face image. A mask is then applied to remove the background and the image is then passed on to a classifier which outputs the gender of the person.

It is interesting to note that this is proof that a 25x25 pixel image contains enough information to determine gender.

The algorithm is also interesting in that it proves that an older and fundamental pattern recognition technique - linear discriminant analysis is just as good as the more trendy Support Vector Machines if used correctly and much more efficient.

Some hard to classify faces: the top row are male images classified as female and the bottom row are female images classified as male.

The research group is also working towards extending the algorithm to recognise facial expressions.

It is suggested that the device can be built to measure television or advertising video audiences by gathering demographic information about spectators. The new device is also useful for conducting market research at shopping centres, stores, banks or any other business using cameras to count people and extract demographic information.

Another application is interactive kiosks with a virtual vendor, as the device automatically extracts information about the user, such as the person's gender, to improve interaction  - unless of course it gets it wrong in which case the interaction may not happen at all!

More Information

http://www.dia.fi.upm.es/~pcr/publications.html