Summer school

Project name

Studying Fish Feeding Practices Using a Computer Vision-Based Individual Fish Identification

Leader

MSc. Mohammad Mehdi Ziaei

Description of the activities of the project

• Studying fish feeding practices is an essential need for sustainable aquaculture. Traditional methods for monitoring the feeding efficiency commonly use group-based metrics. Although individual-based assessments are feasible for some experiments under certain conditions by deploying invasive methods such as PIT tagging, they can not be used for tracking the free-swimming individual fish activity purposes. Moreover, these invasive approaches for individual identification have a deviating impact on fish feeding experiments and measurements. Therefore using non-invasive methods such as visual biometric methods for these tasks can be considered as an alternative solution for measuring and monitoring the feeding experiments for individuals.
• The goal of this summer school project is to develop a prototype computer vision software designed to identify individual fish by leveraging their unique visual features. This approach utilizes video/image analysis powered by deep learning techniques, enabling the accurate and reliable collection of individual-level data for fish feeding analysis.
• During the three-week summer school, students will learn the basics of Python programming while gaining practical experience with AI and computer vision applications. They will also learn about the fish biometric system, focusing on tasks such as labeling data, training models, and integrating various components of the system.
• The final outcome will be a fish biometric software for feeding metric computation and analysis.

Students

• Number of students - 1-2
• high school students students - NO
• university students - YES
• Prepositions (what should the students know): basics of programing