Natural disasters and atmospheric anomalies demand remote monitoring and maintenance of naval objects especially big-size ships. For example, under poor weather conditions, prior knowledge about the ship model and type helps the automatic docking system process to be smooth. Thus, this set aims to classify the type of ships from an image data set of ships.
I have designed a transfer learning-based CNN architecture to classify the data set. The model is evaluated by cohen kappa metric. kappa score = 0.0