Paper Title
Uav-Based Coastal Debris Detection: Analyzing The Effectiveness Of Enhanced Yolo V8 For Improved Aerial Surveillance

Abstract
In the face of the pressing issue of pollution in our oceans, whereas staggering 8 million tons of plastic finds its way into the water each year resulting in a mind-boggling 5.25 trillion pieces of waste, our study presents an innovative method for detecting and identifying shoreline and waterborne garbage. This method is crucial for addressing the problem that sees 70% of debris sink into the marine ecosystem while 15% floats and another 15% ends up on beaches. By utilizing machine learning techniques and unmanned aerial vehicle (UAV) technology, our approach significantly improves both accuracy and speed in detecting garbage. We conducted an evaluation using configurations of the YOLOv8 model along with a specialized dataset captured by UAVs incorporating custom augmentation techniques. Our results demonstrate how this model excels at identifying types of garbage. This groundbreaking research not only advances object detection technology for conservation but also emphasizes the importance of cutting-edge solutions in mitigating the profound impact that plastic pollution has on marine ecosystems. Keywords - YOLOv8, Wise-IoU Modification, UAV Data Analysis, Dataset Augmentation Techniques, Custom Hyperparameters, Object Detection