Inspiration
As humans begin to encroach on animal habitats, keeping track of animal populations is increasingly important as they grow or shrink over time. Learning more about the critters we share a planet with encourages us to be more spatially aware of how close our furry counterparts live among us. AniMaps prioritizes education and real world observations to ultimately promote animal conservation efforts.
What it does
We built an animal tracker app for mobile, with a dashboard for the web, that will be available to the general public. Users can take photos of the animals they see whether on a walk in their neighborhood, or out hiking in the mountains, and the app will use an animal recognition deep learning capability to tell the user what kind of animal it is. If the animal species is considered dangerous, such as a bear or a mountain lion, the app will provide an alert. Once the species is determined, a pop-up will provide information on the species’ background. This app has several uses; it will aid in:
- keeping track of animal species populations with the help of the region’s public
- providing a radar for dangerous animals to be cautious of for a user or user's pets
- allowing conservation groups to better track animal preservation and repopulation efforts
- and many more
The app will also have incentives to get people of all ages outdoors to collect animal sightings with the goal of leveling up and winning awards.
Our app consists of three main functionalities on the mobile end, as well as a web dashboard:
Mobile App
There are three components of our mobile app: viewing the map with animals, adding a animal that you have captured, and searching for the natural range of the creatures in your region. When you land on the home page, you will be able to upload an animal you capture, name it, and drop it into the map. If you don't know what the animal is, you have the opportunity to use our deep learning model to identify the animal. If that stills leave you unsure, you can check the natural ranges of all the animals in your region to feel confident that your animal will indeed exist in your area.
Dashboard
Once you get home, you can hop off of our mobile app and continue to use AniMaps on your computer. Our dashboard provides you with a vast supply of educational resources and all of the observations you and your community have made.
How we built it
We used many ArcGIS product, including ArcGIS Pro, ArcGIS Online, ArcGIS Dashboards, ArcGIS Runtime API for .NET, and ArcGIS API for Python.
Mobile App
We utilized the Android data collection sample in order to have a good start to create our app, we then did a lot of research and tweaking in order to get the app to function like we wanted it to. We used PyTorch to develop a deep learning model and we used the ArcGIS API for Python in order to get the neural network to interface with our ArcGIS Runtime SDK for .NET Android app.
Challenges we ran into
We didn't know how to make the neural network we trained interface with the ArcGIS Runtime SDK for .NET Android and it took us a while to find a solution. As a team, we were also pretty unfamiliar with the ArcGIS Runtime for SDK and we had to learn how to utilize it to make our app. Despite working very hard, we were unable to get 3D models to work in the ArcGIS Runtime for SDK, but we did a lot of research to try to get them to work.
Accomplishments that we're proud of
Successfully developed a mobile app that communicates with a ArcGIS dashboard to add feature points, display pop-ups, filter layers, and view webmaps. We also developed a AI nural network that takes a photo from the map, identifies what animal it is seeing, and send the results back to the user.
What we learned
We learned how to utilize the ArcGIS Runtime SDK for .NET Android and we learned more about neural networks and to make them interface with an ArcGIS Runtime for SDK app. We also learned about how to delegate tasks in order to reach project completion.
What's next for AniMaps
- along with an augmented reality view of a 3D model featuring the animal
- create user accounts so that people can keep track of the animals they have seen
- user testing to get feedback from users and see what they think.
Log in or sign up for Devpost to join the conversation.