I’ve always been skeptical of image recognition apps that try to compensate for a pair of broken eyes. I remember, rather too vividly, a CNIB demonstration of a colour indicator. The thing was outrageously priced, and in any case it really didn’t work. The salesperson didn’t do a very good job of hiding her dismay when it failed, during multiple attempts, to get the colour right—or even close to right. Since then I’ve been, perhaps unfairly, disenchanted with image recognition technology.
an image recognition app called TapTapSee came on the scene and encouraged me to think differently. Sure, it had a few kinks to be worked out, and even today, it’s not always spot on. (During one memorable session, it informed me that a teabag I was photographing said “tips about relationships.”) Despite its occasional mistakes, and its apparent inability to master colour indication, its uses cannot be quantified. It recognizes labels on packaging, articles of clothing, and almost anything else you’d need help to identify. Sometimes, it’s so descriptive that it scares me a little: it once told me that my profile picture included a “woman in a black tank top smiling in a field of yellow flowers.” The detail (and accuracy) was enough to make my jaw drop. It’s worth noting, however, that the magic happens largely because of the efforts of sighted volunteers. Without their insight, the app would be just as clumsy and ineffectual as all the others. Those volunteers, in particular, are what make TapTapSee shine.
It’s still best to label everything and keep my belongings organized. However, it’s nice to know that a clever app like TapTapSee has my back. It has only improved with time, and I can’t wait to see where image recognition technology goes from here.