Automatic Fish Species Classification Using Deep Convolutional Neural Networks

摘要: In this paper, we presented an automated system for identification and classification of fish species. It helps the marine biologists to have greater understanding species their habitats. The proposed model is based on deep convolutional neural networks. uses a reduced version AlexNet comprises four layers two fully connected layers. A comparison against other learning models such as VGGNet. parameters are considered that number fully-connected layers, iterations achieve 100% accuracy training data, batch size dropout layer. results show modified with less has achieved testing 90.48% while original 86.65% over untrained benchmark dataset. inclusion layer enhanced overall performance our model. contain images, memory it also computational complex.