Scopus Indexed Publications

Paper Details


Title
Developing a Classification CNN Model to Classify Different Types of Fish
Author
, Professor Mohammad Shamsul Arefin, PhD,
Email
Abstract

Identifying any fish type can be difficult for people who are not familiar with fish. Implementation of a fish classification machine learning model can become helpful in this scope. The purpose of this paper is to build such a fish classification machine learning model. With this classification model, people will be able to identify the class or type of fish even without much experience with fish. Different types of fish have different nutrition, vitamin, and fat content. Thus, this model can be helpful to ensure better nutrition intake as well. As we have to classify types of fish, we implemented a Convolutional Neural Network (CNN) with Keras along with a modified VGG16 transfer learning model. With the CNN model, the accuracy of our training is 96.67%, and classification accuracy with the modified VGG16 is 97.44%. For validation, with the CNN model, accuracy is 99.92%, and classification accuracy with the VGG16 is 99.76%.

Keywords
"Fish dataset Feature extraction VGG16 Machine learning Transfer learning Image augmentation Image classification"
Journal or Conference Name
Lecture Notes in Networks and Systems
Publication Year
2023
Indexing
scopus