CROWDPET: “Have you seen Smudge”?
In cities around the world, lamp-posts and park railings are regularly festooned with piteous messages – often written in heart-rendingly childish script – attached to faded photographs of cherished but lost family pets. Cats, dogs, escaped parrots, even the occasional anaconda. Missing pet flyers are a global tug on humanity’s tender heartstrings.
Thanks to children’s classics such as “Lady and the Tramp,” the theme of the lost dog reunited with its owners has a powerful effect on our collective imaginations. There’s even a movie about “Missy the Missing Cat“. This is a global need that provides a practical application for some serious science, too.
For today there’s an incipient tool to harness the power of the internet to widen the search for all those pooches gone walkabout. CrowdPet, a smartphone app developed by SciPet, a Brazilian startup incubated at the University of Campinas (UNICAMP), uses facial recognition techniques to establish a database of domestic animals.
The technology developed almost by accident, as a consequence of a census of animals reached by vaccination campaigns carried out in July 2017 by the Centre for Animal Disease Control in the city of Vinhedo, São Paulo State. But the app’s inventors found they had couple be tapping into a potentially huge private sector market.
CrowdPet is designed to combine two data sources: photographs of animals registered by their owners and photographs of lost animals sighted in the street or brought in to rescue centres by volunteers. “The app establishes a match between two images using computational visual recognition, and pinpoints by geolocation the place where the picture of the lost animal was taken,” says Fabio Rogério Piva, who is the CEO of SciPet and has a PhD in computer science from UNICAMP.
CrowPet’s business proposition is to harness social media to get animal owners to photograph and upload images of (extended) family pets. Petshops, vet offices and those with a commercial interest in animal welfare can participate. At the same time local municipal administrations and animal welfare bodies such as NGOS are encouraged to photograph all lost animals coming into pounds so they can speedily be reunited with their owners. Cities are now considered the main partners for the project. “CrowdPet can be offered free of charge to the public through local government,” Piva explains.
The research project, conducted by students with grants under the supervision of Fernanda Andaló, SciPet’s CTO, was selected among finalists for the 2017 Inova UNICAMP Prize. The next step is even more complex, requiring accurate identification of each animal. The project also got financing from Brazil’s largest regional research funding institute, FAPESP.
This institution’s Innovative Research in Small Business Program (PIPE) paid for a study in which its feasibility was analysed. SciPet completed a prototype capable of distinguishing cats and dogs from all other images with a 99% success rate. “Even if a user photographs people or objects, the system will register only photos of animals,” Piva explains.
The first version of SciPet’s business plan for CrowdPet chose people who wanted to find missing pets, connected in a kind of “social network”, as its sole target clientele. However, before completing Stage 1 of PIPE, SciPet was selected to take part in the Fourth Edition of FAPESP’s Training Program for High-Tech Entrepreneurs in 2017.
SciPet was founded in 2016. According to Piva, the firm’s first studies were based on a project called “Where is my puppy? Retrieving lost dogs by facial features”, conducted by researchers at UNICAMP’s Reasoning for Complex Data Laboratory (ReCoD Lab), with Professor Eduardo Valle as principal investigator. Valle is currently working with SciPet.
“This was one of the first attempts to apply biometry to animals,” Piva recalls. “It showed that human facial biometry methods aren’t very effective to identify dogs. Using a method specially developed for animals, they achieved 89% accuracy, equivalent to the success rate obtained by a human observer who specializes in dogs.”
Right now the system’s biggest problem is that a well-groomed pooch in a happy home looks very different to a bedraggled and starving stray animal who’s been living on the street. It’s hard for recognition software to connect such radically different snapshots.
To read more about CrowdPet you can click on this link for a report by Brazilian journalist Suzel Tunes.