To subscribe to the ORE seminar mailing list, click here.
For a (working draft) list of upcoming seminars, click here.
If you are interested in giving an ORE seminar, please contact us at nosal [at] hawaii [dot] edu.
- This event has passed.
Seminar: How can AI scale up coral reef monitoring? Semantic mapping of coral reefs from videos using deep learning.
Jonathan Sauder
PhD student at the École Polytechnique Fédérale de Lausanne (EPFL) in Lausanne, Switzerland working on the intersection of 3D computer vision, machine learning and coral reefs
*In-person (HOLM243) & via Zoom*
Zoom Meeting Link
Meeting ID: 890 6867 4318
Passcode: ORE
In light of the existential threat to coral reefs worldwide due to human activity, there is an urgent need for innovative monitoring strategies that are efficient, standardized, scalable, and economical. Mapping coral reefs from underwater videos in real time using consumer hardware could enable the next generation of coral monitoring tools. However, coral reefs present challenging conditions for 3D computer vision algorithms, often causing established systems to fail when processing casual videos. In contrast, 3D mapping systems based on deep learning offer increased robustness, allowing transects to be mapped within minutes and enabling seamless integration with neural network–based semantic segmentation. This technology underpins the surveying methodology DeepReefMap, which can substantially reduce the cost of coral reef surveys. DeepReefMap is now implemented in four countries and is used to analyze a steady stream of monitoring data from dozens of reef sites, with preliminary results capturing the effects of the 2023–2024 El Niño on reefs in the region.