Facial Recognition – A Better Solution for Television Ratings?

January 18, 2013

By Bill Duggan

We all watch television.  Now some televisions have the capability to watch us back.  Google the term “Facial Recognition” and you’ll see that there are 5.4 million search results!  A facial recognition system is a computer application for automatically identifying or verifying a person from a digital image or a video frame from a video source.

Companies including LG, Samsung, Toshiba, and Hisense (China’s leading TV manufacturer), have televisions with facial recognition technology.  A built-in camera recognizes the face(s) of the individual(s) in the room and then personalizes the experience to each recognized family member, switching to different preferences and home screens based on which member of a family sits down to watch.  If you have visitors over, those faces could be logged as well.

Which brings me to television ratings.  Clearly, both buyers and sellers want the most accurate ratings information possible.  That is undeniable – the more accurate, the better.  The current People Meter requires a viewer to “punch in” to a device hooked up to their television when watching TV and then “punch out” when leaving the room (even for a bathroom or refrigerator break).  Nielsen introduced the People Meter nationally in 1987, and while that was revolutionary at the time (especially compared to manual diaries), a lot has change in those sixteen years.

Facial recognition has interesting applications for television audience measurement to give a more accurate count of viewers in front of the TV, while not relying on the compliance of viewers to punch in or punch out.  There are privacy concerns, of course.  But facial recognition is an idea worth exploring for television ratings.  I’ve heard that there is some work already being done here, and encourage the sharing and acceleration of that work.


You must be logged in to submit a comment.