Summer school

Project name

Fish Fin Damage Detection and Classification Using AI

Leader

MSc. Stephany Osei

Description of the activities of the project

• Fish fin morphology is an important indicator of fish health and welfare, as fin damage may reflect stress, aggressive behaviour, or unsuitable environmental conditions. Such damage can negatively affect swimming performance, feeding behaviour, and overall fitness, making its timely detection essential, particularly in captivity and aquaculture systems.
• This project focuses on the use of image analysis and artificial intelligence for the detection and classification of damaged and undamaged fish fins.
• Students will work with a large dataset of salmon and trout fin images and participate in the annotation process to create labelled training data. The annotated dataset will be used to train convolutional neural network–based object detection models.
• The project aims at exploring and comparing different ML architectures to evaluate learning performance, training efficiency, and ease of use.
• During the project, students will gain basic experience in data acquisition, image processing, annotation, machine learning, and neural network training, while gaining insight into the application of modern AI tools for biological image analysis.

Students

• Number of students - 1-2
• high school students students - NO
• university students - YES
• Prepositions (what should the students know): Basic algebra, basic algorithmization, basic statistics.