Training;Sea surface;Sensitivity;Computational modeling;Autonomous aerial vehicles;Search problems;Convolutional neural networks;Unmanned Aerial Vehicles;Search and Rescue;Image classification;Convolutional Neural Networks;Decision making