FruitPunch AI is hosting AI for Pelicans, together with Rewildling Europe and the Romanian Ornithological Society, to help researchers save valuable time by automating the detection of pelican nests.
In this Challenge, we will train machine learning models on aerial photographs of a breeding pelican colony in Romania's Danube Delta to count the number of nesting pairs and develop a user-friendly interface for biologists.
This is a perfect Challenge to develop or show off your computer vision skills and build your portfolio in AI for Wildlife! FruitPunch AI organizes domain and tech masterclasses and you will be working in a team with people from all kinds of professional and cultural backgrounds.
𝐈𝐧 𝐭𝐡𝐢𝐬 𝐜𝐡𝐚𝐥𝐥𝐞𝐧𝐠𝐞, 𝐰𝐞 𝐰𝐢𝐥𝐥:
- Extract label-pixel coordinates from large aerial imagery
- Estimate Bounding boxes for the pixel coordinates
- Build CNNs
- Train object detection models
- Create a GUI for researchers to check and adjust the predicted labels and pelican count
The Challenge takes 10 weeks. By joining for 8h/w, it's a great opportunity to
- get started with ML in the Wildlife domain
- get your hands dirty with real data, working for real stakeholders
- learn from enthusiasts and experts from all over the globe, both peer 2 peer and in Masterclasses
- build a network in AI for conservation
Everyone from biologist to computer scientist can join as long as you have basic knowledge of Python programming and machine learning.
You can apply straight away through our platform
Learn more about FruitPunch AI
Add the first post in this thread.