An open source textbook for facilitating undergraduate education in marine computer vision

As careers and internships in marine imaging continue to grow, there is a notable lack of customized educational resources for onboarding and skill-building in this specialized field, particularly in marine image data literacy. To address this gap, I present a pilot course titled "Computer Vision across the Marine Sciences," which will be taught as a special topics course during the winter quarter of 2025 and made available online indefinitely for free. The course is accompanied by an open-source interactive textbook and lab manual, available at oceancv.org, which provides hands-on learning materials. This course aims to equip students with best practices and current research in marine imaging techniques, alongside intermediate Python programming skills. Our instructional approach is grounded in constructivist learning theory, emphasizing active, self-guided exploration. Key components of the course include the use of Google CoLabs to facilitate interactive, cloud-based learning environments where students can engage with Python syntax and functions relevant to marine imaging. Through a combination of flipped classroom structures, synchronous activities infused with active learning, and individualized final projects, students will be encouraged to apply their learning in practical, research-driven contexts. The course design prioritizes accessibility, ensuring that learners with varying levels of prior experience can achieve similar success through higher engagement with course resources. By analyzing data from surveys, student work, assessments, and focus group feedback, I aim to continuously refine the course to better support self-guided scientific inquiry and skill acquisition. I believe that this approach to teaching computer vision and Python within the context of marine imaging will be broadly applicable and beneficial across the earth sciences and other scientific domains.  

Presentation Preference: Either

Primary Presenter: Atticus Carter, University of Washington (attcart@uw.edu)

Authors:

Atticus Carter, University of Washington  (attcart@uw.edu)