We have invites, click through to get yours!
Earlier this year, a company called Face.com brought facial recognition technology to Facebook by way of an application called Photo Finder which scanned through untagged photos and identified the people within them. Now, using the same facial recognition algorithms that made Photo Finder possible, the company is introducing Photo Tagger, an app which scans through select online albums to automate the tagging process.
The two Face.com Facebook applications are very similar in nature. They both use the company's facial recognition technology to match people with their pictures by way of a special algorithm called the "hybrid descriptor-based funneled" model. To the layperson, though, all that matters is that the technology makes facial recognition possible even in "everyday" photos - meaning photos taken from different angles, out-of-focus shots, photos in low lighting or those in which people are making odd facial expressions, etc.
About Photo Tagger
But where Photo Finder focuses on discovery, Photo Tagger focuses on productivity. With the new app, you can choose the albums to scan - whether your own or those belonging to your friends - and the app will process the photos they contain. Photo Tagger will batch the people it finds into groups and will then suggest tags for them. Once you confirm the tags, they're automatically pushed to Facebook where the people tagged are notified, just as if the process had been done manually.
Facebook users who upload a lot of photos will find an application like this extremely useful as it dramatically cuts down on the time it takes to tag images. Instead of having to go through each photo one-by-one, you can simply confirm the tags the application suggests. And surprisingly, it doesn't make a lot of mistakes. The facial recognition technology employed by both of these applications is incredibly accurate. In our tests, the most common mistake it made was to identify a picture of a friend's child as the friend themselves - a problem that has a lot to do with how facial features between related family members are so alike. But for the most part, the app identified photos correctly.
The Photo Finder app has already produced amazing results. It has scanned more than 1.5 billion photos so far and has identified more than 2.3 million faces. What's even more shocking about these numbers is that the app is still in closed alpha. Imagine how many faces it will scan when it opens up!
Get in on the Alpha
Photo Tagger will also launch in closed alpha but will probably be closed for a shorter period of time than the Photo Finder application. The reason for this is because Photo Tagger scans albums on the fly as opposed to scanning an entire network of inter-connected users and their photos like Finder does - a number which can be on average over a hundred thousand photos. Since that causes a bigger impact on the service's back-end than Photo Tagger, it will remain closed for longer while the company works out the kinks.
No matter, because out of the two applications, Facebook users will likely find Photo Tagger the more useful of the two since it offers a more direct purpose: tag these photos fast!
Photo Tagger launches today in private alpha. If you want in, you can click this link to add the app to Facebook right away. Only the first 100 ReadWriteWeb visitors who do so will be given access. Good luck!