Not too long ago, he and his wife were in a coffee shop and spotted a lost dog poster. It brought back memories of when his family dog went missing -- he was distraught; his kids were in tears. And that spurred an idea: Why not use facial-recognition technology to help track down Fido?
Every dog owner reading this is saying, "That's genius!" But as Polimeno quickly learned, it's not that simple. The feature that arguably makes dogs so darn adorable also makes facial recognition incredibly difficult -- their fur.
Polimeno started calling the major players in facial-recognition technology and finally found a willing partner in the Software Development Center at the University of Utah. He funded the research while the center cracked the nut of facial recognition for our furry four-legged friends. The programmers ended up developing an algorithm called Pet Match that uses machine learning and computer vision to pick up on key features of dogs that differentiate them from the rest of the puppy population.
Watch the video to see exactly how the app works and the clever built-in photo feature that helps you capture your pup's best side.
Finding Rover recently hit the Apple App Store and is free to download. Android and Web versions will be available in a few months, so you can use the technology even if you don't own a smartphone.
Besides helping dogs and their families reunite, Finding Rover serves as a type of brag book. You can share photos of your dogs with your social circle, or your "pack," and a tips-and-tricks section offers a place to post advice on everything from how to travel with your dog to how to train them.
And for all the cat lovers out there, don't accuse Polimeno of being biased toward pooches. Finding Kitty is in the works.