IsVoNet8: A Proposed Deep Learning Model for Classification of Some Fish Species

In the classification of fish, both knowledge and great effort are required to determine the characteristics of fish. Traditionally, however, manual classification of extrinsic characteristics of different fish species has been a difficult and time-consuming process due to their close resemblance to each other. Re-cently, deep learning methods used in the light of developments in the field of computer vision have facilitated the training of fish image classification models and the recognition of various fish species. In this study, a new convolutional neural network model classifying 8 different belonging to 6 families (Mulli-dae, Sparidae, Carangidae, Serranidae, Clupeidae, Salmonidae) fish species using deep learning methods was proposed. The species include Clupeonella cultriventris N., Sparus aurata L., Trachurus trachurus L., Mullus barbatus L., Pagrus major T & S., Dicentrarchus labrax L., Mullus surmuletus L. and Oncorhynchus mykiss W. The proposed model (IsVoNet8) is compared with the ResNet50, ResNet101 and VGG16 models. The success accuracies obtained as a result of the comparison are respectively; 98.62% in the IsVoNet8, 91.37% in the ResNet50 model, 86.12% in the ResNet101 model and 97.75% in the VGG16 model. However, it was obtained that the loss rates of ResNet50 0.3646, ResNet101 0.5811, VGG16 0.0696 and with the IsVoNet 0.0568. As a result, it has been observed that the IsVoNet classifies marine fish, which is widely consumed in Türkiye. © 2023 Kaya et al.