Wildlife trafficking happens everyday across the country. Stopping the trafficking of protected species can be very complicated, it can be very difficult for law enforcers and the public to identify illegal animals for many reasons. With hundreds of protected species in the Philippines, it is nearly impossible to identify each one of them from memory.
*It is our team’s goal to help put an end to illegal wildlife trade. *
One of the hardest things about stopping wildlife trafficking is accurately identifying which species are protected, and shouldn’t be in the market. We realised that making that job easier for people will be one of the best and practical ways to help stop the illegal wildlife trade.
What it does
Wildify is an AI-driven mobile app that helps both law enforcement and the public easily and accurately identify endangered animals. By just taking a picture of an animal you have seen, Wildify can identify that animal and show its scientific name, common name, description and the conservation status or classification of the animal you have identified. Users can also use Wildify to instantly report any wildlife crimes to the authorities.
We have also added an interactive educational feature which brings users a new way of seeing and learning about Wildlife in Augmented Reality. Wildify allows users to scan printed animal cards and witness how the animals come alive right before your eyes. This gives users an unforgettable experience and will leave an impact to the children and adults alike. Wildify gives users both interesting and informative facts about endangered and highly trafficked animals.
How We built it
Wildify is built with Swift and can run on iOS devices. The machine learning model was trained with CreateML to identify endangered animals. Currently we have 6 endangered animals in our model and we downloaded 50 images for each animal as training data, and 10 images for the testing data. The larger the data set, the higher the accuracy. The augmented reality feature incorporates SceneKit to render the 3D animal objects. Animal cards were printed to serve as trigger/reference images for the AR environment.
Challenges We ran into
Our machine learning model is not that accurate especially when the lighting is not good. We need to train our model to a larger data set and we can do that by partnering with the Department of Environment and Natural Resources to gather more animal images for training.
Accomplishments that We're proud of
Wildify will be the first in the country that will identify endangered species by just capturing the animal’s image with the use of Machine Learning and that is something our team will be proud of.
What's next for Wildify
We want to make it available for download on the App Store, but we need to optimize the app further in terms of accuracy and add more 3D assets for our AR feature. We will also try to develop an Android version of the app so that the app can be enjoyed on both mobile platforms. We want to market this app on popular social media platforms, we strongly believe that a mobile application is a practical solution to help end wildlife trafficking. We all share the responsibility for protecting our country’s wildlife before it’s too late. If we all work together, there is still hope for these beautiful creatures.