Tuesday, January 13, 2009
Currently, I'm engaged in an activity I should have done months ago: Compile more detailed profiles for the characters in my novel (including photos). Being a visual learner, I need to see photographs of my characters in order to bring them to life in my fiction. For "Plainfield" I realized that some of the characters were thin (even the protagonist) because I didn't have a good mental picture of them. Therefore, I scoured through online yearbooks from the period (1950s), Google Images, and family photo albums on Flickr looking for my characters. I had no real criteria other than trusting my gut to match my mental image with a face online. I spent hours doing this, thinking up different (and often random) search criteria. Sometimes I would just search the character's name and find all the people named Walter. Other times I searched a character's profession, such as "teacher," "sheriff," and so on. My thinking is that on some metaphysical level, a teacher out there will match my teacher character (people of a certain physical nature are attracted to certain jobs).
The problem with searching this way is that often I would get images tagged with my keyword, but that didn't show anyone--just a pic of a place, a car, a building, etc. After a little digging, I discovered an article on Ars Technica detailing a cool trick to take advantage of Google Image's facial recognition feature. Basically, do your image search, then append "&imgtype=face" to the end of the result URL. Hit Enter, and the results will be reloaded with only those that contain faces (as in the screenshot of this post). Click here for the full article. This is a pretty cool little trick for fiction writers who want to create in-depth character profiles that include photos (and not those of people you know).