Trello board

cheerios015I may be one of the few people that enjoy watching tv commercials. I have even found myself sticking around for youtube video ads. Besides being intrigued by the craft that is involved creating these messages, I find them to be 30 second studies on American culture. In them are hidden trends about how we collectively look at gender, sexual orientation, age, and race.

(Take a look at one of my favorite series of commercials from the mid 90s: can you spot where they got it right?)

Recently, cheerios, a brand that is synonymous with breakfast, america, good heart health, and family, chose to portray a mixed race family in their home  enjoying cheerios. The comment section of sites showing the ad exploded. Some americans did not take well to that image of a typical family, even though it is a reality of the american family.

Brands and companies should be applauded for portraying the diversity of the american family but this very small event, although coming more common place, is not the rule.

I am fascinated by this transition from ads the past that could be seen as racist, homophobic, misogynist, sexist and chauvinist.  I want to find as many american advertisements over the past few decades and analyze one aspect to them: whose faces appear in them. Using facial recognition software through the openCV library, I want to find, collect, and portray the faces of brand building and capitalism. I also want to create facial composites of what the advertising community believed would be the best received representantive of their products.

edit: possible corpus? the award winners of advertisements for the past 5 decades.