AI for Turtles
Develop computer vision software that can recognise and distinguish individual turtles through automated identification.
Premium participation: Get 1-on-1 Mentoring & Guidance
Try a free mentoring sessionHelp the conservationists by developing AI models and monitoring software to protect the turtles
In the Caribbean, conservationists regularly monitor turtle populations and have collected extensive data on these creatures, including individual labeling. The goal of this challenge is to use AI to automate the population monitoring process.
The current method of tagging turtles for monitoring is intrusive, time-consuming, and expensive. To address this issue, automated recognition software using photo and video data can enhance monitoring efficiency and assist organizations dedicated to protecting this endangered species. Sea turtles have unique facial scales similar to our fingerprints, and AI can identify these characteristics to associate them with a particular turtle.
This Challenge focuses on data collected in the Caribbean, particularly in Bonaire, which is home to the loggerhead turtle, the green turtle, and the hawksbill turtle - all listed as either endangered or critically endangered. Bonaire has taken significant steps towards sea turtle conservation, including establishing protected areas where nesting sites are marked and monitored to prevent disturbance. These areas are patrolled to prevent poaching and other human activities that could harm the turtles or their nests.
Responsible tourism is also recognized as an essential aspect of sea turtle conservation in Bonaire, with local organizations educating visitors about the fragile nature of sea turtle populations and implementing snorkeling and diving guidelines to ensure that these activities do not disturb the turtles or their habitats.
To help the conservationists we will be developing AI models and monitoring software to protect the turtles, and you can help!
Info session
Our challenge partner
In this challenge, we will
- Research Siamese and Triplet networks to perform turtle identification
- Research other state-of-the-art techniques like ArcFace
- Build data pipelines for processing train and test sets as well as training models
- Create an easy-to-use GUI (Graphical user interface) in collaboration with the researchers.
Who are we looking for?
AI/ML enthusiasts, Software developers, Biologists with a knack for AI, and computer scientists.
You will collaborate with a diverse team of up to 50 international participants in subteams. You can join as a contributor (8-12 hours per week commitment for 10 weeks) or coach (2-4 hours per week, only for experienced ML professionals)
We’ll organize masterclasses on marine biology and animal identification
Timeline
Application Deadline: 11 October 2023
Challenge Kick-off: 18 October 2023
Midterm Presentations: 15 November 2023
Final Presentations: 10 January 2023