I'm sitting on a stool in a plain white room at the Palo Alto Research Center, checking out my new earrings via a small desktop mirror. They're a big dangly pair, each with a white porcelain rose and a black stone hanging underneath. On top of the mirror is a webcam attached to the computer on its right &mdash it's recording my every move and sensing the angles of my head and the closeness of my face as I check myself out. After a couple of minutes of striking poses, I click on the mouse to pause the camera and take the earrings off. I put my old pair back on &mdash a petite gold and silver lotus root that I bought in Japan several months ago &mdash and press play. Two screens pop up, one of me now and one of me just a few minutes ago with the previous pair. It's like I'm seeing double &mdash every time I move my head in real time, the me from a few minutes ago moves her head the exact same way. The flower earrings, I notice, look a lot better from the side, but I like the way the lotus roots dangle when I'm looking straight ahead.
By streaming video taken by the camera through their spatially oriented machine learning software, PARC researchers have figured out how to give people like me a real-time interactive comparison shopping experience. The responsive mirror system, which comes in both a desktop and full-length version, displays previously worn outfits on a second "mirror" &mdash the playback of a movie taken by a webcam on the ceiling that locates you spatially within the frame and then finds the same angled shot from the previous clip. The technology hasn't hit retailers yet, but PARC researchers are hoping to implement it in dressing rooms soon.
The idea of using digital images to compare outfits, for me at least, dates back to 1995, when the movie Clueless came out. In the opening scene, Cher (Alicia Silverstone) is picking out an outfit to wear to school. "Actually, my life is way normal for a teenage girl," she says. "I wake up, I brush my teeth, and I pick out my school clothes!" The camera then zooms in on an old school computer monitor showing a program that helps her pick out an outfit. Cher scrolls through all the items in her closet, picks a few possible outfits, and then the software juxtaposes the best look onto a photograph of herself. Like many teenage girls in the 90s, I was dying to own that cool software that let her "try on" multiple outfits at the same time and alerted her of mismatches. Imagine how much insecurity and embarrassment that would have helped me avoid in high school, to know that my outfit was perfect every day!
Of course, there are some obstacles to this becoming mainstream &mdash people aren't always comfortable with having a camera in the dressing room (although this has become fairly common as an anti-theft mechanism), and in this economy I'm not sure that retailers are willing to pay extra to have fancy technology integrated into their existing analog dressing room system. But I can think of some instances in which it would really make sense. Take this real-life example: in February, I spent a full hour in front of a handheld mirror at an eyeglasses store in Harajuku trying on three glasses frames over and over. I am not an avid shopper but I wanted a solid pair of glasses that could go with any outfit, and I just could not figure out which one looked better! I even had the saleswoman take pictures of me in each one so I could study them side by side on my G10. I'm willing to bet that this high-traffic retailer would be willing to invest in the desktop responsive mirror so that the salespeople don't have to spend so much time babysitting indecisive customers like me. Plus, I think it would be a lot of fun for me. Shopping, after all, should be fun. It shouldn't feel like hard work.
About Lisa Katayama
I'm a contributing editor here at Boing Boing. I also have a blog (TokyoMango), a book (Urawaza), and I freelance for Wired, Make, the NY Times Magazine, PRI's Studio360, etc. I'm @tokyomango on Twitter.