VentureBeat

Fresh from its launch a week ago, the new Chinese social network City!N is launching today one of the features that it hopes will differentiate it: face detection.

The tool draws a box around any faces it finds in pictures uploaded to the site. Users can then tag the face by typing identifying information into a database. They can then automatically share a photo with that friend’s face, since the social network can dispatch an alert to the friend saying a photo of them has been uploaded, said Simon Chan, co-founder of City!N.

“We are showing the world what an intelligent social networking service means,” Chan said.

At some point, users will be able to search for that same facial image across the network. They can thus find all of the pictures associated with that face on the site. The tool raises privacy concerns, since many users might not want their wild and crazy pictures found by distant friends. But Chan says that users are free to set their own photos to “private” settings so that they can’t be shared or discovered.

This new kind of social discovery tool is what Chan says will set the company apart. Eventually, the tool will be used to find people who might be a good match for the user because they meet certain facial characteristics.

There are other face-detection search engines available. Polar Rose of Sweden has a beta version of its face detection search engine. Polar Rose tries to avoid heavy computational requirements with a unique design. It takes the faces in pictures and converts the characteristics in them into 3-D animations. Those 3-D constructions are much easier to use in recognition tasks, since there is far less data to be computed, in contrast to the data associated with processing a photo of a face.

I’ve used the Polar Rose search engine with good results. VentureBeat wrote about a number of different companies pursuing the same face-detection technology. I haven’t been able to evaluate City!N’s technology. But if it succeeds, you can bet that Facebook, MySpace and others won’t be far behind with applications of their own.

As far as I can tell, they don’t yet have it. But the applications for face detection will run far and wide. My colleague MG Siegler points out Eyealike has an application where it uses face detection to check for copyrighted video piracy. I saw Eyealike at DEMO, and they showed how they could recognize patterns of motion and compare that against a database of frequently pirated films.

Banks use face detection technology from companies such as 3VR to detect known check-fraud criminals when they walk into banks.

The company’s web site is CityIN.com, without the exclamation point that is part of its formal name. City!N is incorporated in Hong Kong and its offices are in Guanzhou in southern China. It has 11 employees and is angel funded.