WildTrack to use deep learning to save endangered wild species

June 7, 2017, 8 p.m. By: Pranjal Kumar


WildTrack is a non-profit organization that is devoted to saving endangered species. On the World Environment day, they have launched a program called ConservationFIT. The term FIT stands from “footprint identification technology”.

According to Zoe Jewell, one of the co-founders,” Finding any wild animals much easier if we use if footprint for identification. This technology not only tracks the animal but also gives much vital information”.

The program will use the crowdsource photos of animal footprints. It will then use these photos to build an algorithm that can identify the individual, species, sex, age class of the animal who made them.

Zoe Jewell said,” The current technology uses radio collaring and dehorning of black rhino as the protection methods. But these methods have not succeeded as the collar were failing regularly. There were other problems like the constant immobilization of females for re-collaring. It reduced their birth cycles from having one calf every three years to one in every 10 years. “

Sky Alibhai is the other co-founder of WildTrack. He said,” It’s important to preserve these animals and if modern technology can help us to that efficiently then why not use it.”

WildTrack has so far developed the FIT algorithm for 15 species. Conversation FIT will help to build the algorithm faster using the images uploaded from field biologist, trackers, and scientist armed with smartphones. The technology’s works with SAS using NVIDIA GPUs for the deep learning methods.

The company aims to take advantage of 8 million visits by recreational visitors make to the protected areas around the world. They will prompt them to take images of the footprint on their smartphone which will help FIT to identify the species distribution and that animal.

Image Source: Nat geo