The Problem: Watching TV does not allow for me to get more information about what I am viewing. For example, a frequent problem is being able to identify who a person is on the screen.

The Solution: Combine facial recognition/character recognition with publicly available databases to provide more information to the user.

Frequently I'll find myself watching a TV program and wondering who a particular actor is and what else she has been in. As a viewer I would like to pause the current program and select the actor I am interested in finding more about. This selection should bring up the actors name and a list of other appearances. If I am interested in viewing another program I should be able to scroll thru that list and select one of those titles.

Beyond TV shows and movies let's apply this to sports shows. Want to find out the latest stats about player X? Click and you shall see. Want to see past highlights, news pieces, or TV commercials of the sports player? It is all right there at the click of a button.

The above two examples, using actors and sports figures, should be relatively quick to do. They both have lots of photos available for scanning and, in addition, the TV program can be flagged with meta data about who is in the program. Data on actors can be had from the IMDB and for there are numerous databases of sports statistics.

What about TV news shows? News shows are littered with pundits that little information, other than name and title, is provided about who they are. Since the pundits name is generally added to the lower third of the screen, character recognition software could be used to "read" the screen. Or, if the program is closed captioned, that text can be used to feed the query.

More information to the viewer is key. This will enrich the experience for the viewer yet be invisible to the passive viewer.