Advancing Marine Species Detection: A Comparative Analysis of YOLOv5 and YOLOv8 in Challenging Underwater Environment

This work offers a thorough comparison of the YOLOv5 and YOLOv8 object detection algorithms with a focus on the identification and categorization of ten different marine species. To address the urgent demand for sophisticated computer vision solutions in marine biology, we assess these models' performance, paying particular attention to their accuracy and effectiveness in difficult underwater settings. The chosen marine species are diverse in terms of size, shape, and color, providing a strong baseline against which to evaluate the models' performance. The investigation and results provide light on the advantages and disadvantages of the models by carefully examining detection metrics, false positives, and false negatives. The study offers helpful advice for scholars and practitioners working in marine biology and other related fields by providing concrete suggestions for model selection based on observed performance.