"This approachable and fun app raises awareness around pet adoption in a way that's guided by an age-old meme." —juror Michael Potts
"It's rare that you see someone use complicated computer vision technology to create something so seemingly simple. It makes technology feel a little bit like magic again." —juror Keith Butters
Overview: Over 100,000 dogs are needlessly euthanized every year in Australia. To overcome people's hesitation to visit shelters, this app brings the dogs to them; to change people's perceptions that shelter dogs are "damaged" it shows that they're just as loving, loyal and eager to please as any dog; and, instead of relying on the one or two shelters within a local area, it brings every available dog in Australia into the palms of people's hands. A canine analysis engine captures the facial features of each dog. When a human face is uploaded the app scans facial features and analyzes and compares them against every dog in the PetRescue Australia database and finds the perfect dog match.
• Each week, over 500 new dogs from 716 rescue groups are uploaded to the live PetRescue Australia database.
• Users whistle to initiate a call to the adoption center and shake the phone to make a donation.
• A match rate of no less than 93 percent is returned every time.
Comments by Dean Hamilton:
What was the response to this project? "It was the number 1 app in the Australia iTunes store for two weeks; there have been over 1,000 downloads every day since its launch and there are 5.8 million Facebook users; there's been a 36 percent increase in dogs adopted (over 2,200 dogs every month) and 3,373 dogs were rehomed during the eight weeks after the app launched—all of it without spending a single cent on advertising."
How did this project compare with others you've worked on in the past? "The concept is based on well-known folklore that many dog owners look like their dogs; the reason it worked so well is that it provided an interesting and fun way to talk about a serious issue. Rather than bombard people with charity messages, we approached the topic in a disruptive and memorable way. As far as we know, prior to this, facial recognition was limited to humans, so we had no idea if we could actually achieve what we wanted to do but we thought it could be interesting to try."
Did you learn anything new during the process? "We learned a lot. Primarily that the better the profile photo of the dog, the better the match—and the greater the chance of each dog being adopted. With this in mind, we educated the shelters as to best practice when photographing the dogs to help make their job easier and the results better."
