At video chat network Rounds.com, for example, face detection analytics comes into play in a number of ways, said Natasha Shine-Zirkel, marketing director for the community in a recent phone interview. In one use, Rounds.com, a “hangout” platform allowing Facebook members to video chat with friends, play games, or listen to music together, introduces new friends using an algorithm to match people by profile data and face detection analytics to match them by age.
Rounds.com also uses face detection analytics to flag members accessing age-inappropriate parts of the site or those showing inappropriate images in the video chat section, Shine-Zirkel said.
In another example, dating community FindYourFaceMate.com uses the same detection analytics to establish matches based on similarity of appearance -- its unique premise, said Gil Hirsch, CEO and co-founder of Face.com, the company providing face analytics for this company as well as Rounds.com.
Face.com provides free APIs that developers can use to create a variety of applications for face detection and recognition software. It offers enhanced services for a fee, said Hirsch in a recent Skype interview.
In the case of sites like Rounds.com and FindYourFaceMate.com, the Face.com detection software looks at face attributes in photos but does not deal with identity, Hirsch said. For example, the face detection software identifies characteristics like profile, orientation, points of interest (eyes, nose, etc.), gender, mood, age, and whether the face might be smiling, wearing glasses, etc.
By contrast, face recognition analyzes images using either Facebook connections or independent databases to identify users.
With the simplest application, you can detect when a photo of you or one of your connections is posted somewhere within your Facebook network. Another app, Celebrityfindr, locates photos of celebrities on Twitter using Google photos for recognition purposes.
Some companies are putting facial recognition to practical use, too. For example, employees at one construction site check in and check out at a construction site without having to punch a time clock, Hirsch said. The app requires a database of employee names matched with photos.
How might your organization put face detection or recognition software to use? Leave your comments below.