New iPhone app identifies friends and tags photos on FB

A new iPhone app that automatically recognises and tags friends on Facebook has been unveiled.

London: A new iPhone app that automatically recognises and tags friends on Facebook has been unveiled.
The app called KLiK performs real-time facial recognition to automatically identify and tag friends in photos.

The app relies on a connection to Facebook, only friends in a user’s network can be identified - to start with.

The app, launched by the facial recognition technology platform, does also have a ‘learn’ mode that allows it to tag people from outside Facebook, a newspaper reported.

“It’s not like you can point this at someone on the street and make it work,” said Gil Hirsch, the CEO of

The app mimics a built-in function on Samsung’s much-hyped Galaxy S3, due to launch on May 29.

It can also add Instagram-style filters and hand-chosen fonts layered over pictures.

It allows users to teach the app who someone is by pointing the camera at them and manually entering their name.

“It’s all private and on your device only,” explained Hirsch, adding that the person will then be tagged automatically.

“It’s our most recent evolution of both the platform and the consumer product that we’re offering.

“We noticed that at parties or events there were many photos being taken but only a few were actually getting tagged,” he cited.

By connecting with Facebook, the app scans friends’ photos to develop a facial profile of everyone in a user’s network. The app identifies people by matching faces in photos taken with, or uploaded to, the app to these profiles.

Users can also apply Instagram-style filters and share photos via Facebook, Twitter or email.

Although the app is only able to identify Facebook friends, or people entered manually, some critics are concerned about privacy issues.

“This system has been engineered from the get-go to preserve privacy and also deliver a social fun value and nothing creepy,” Hirsch said.

The company claims KLiK is approximately 90 percent accurate.