Paper Title
Using Faster R-CNN to Detect and Identify Offshore Fisheries

As the Artificial Intelligence (AI) has continuously set an overwhelming pace for rapid growth and more mature, new applications of AI-based activities, have now become a part of people’s daily life as well as in business landscapes. The main purposes of this study were twofold. First, we systemically identify the type of fish and detect of the coordinates of the object by using Faster R-CNN method, moon fish, tuna, ghost head knife, and shark served as examples. Second, from the viewpoint of comparison, the study compared the Faster R-CNN with SSD (300x300) method in terms of precision. The experimental results of this study revealed that Faster R-CNN with the accuracy of 95.1%, higher than SSD (300x300) 90.4%. It is hope that the method used in this study could be adopted as one of the methods for identifying fish species in the electronic fish monitoring system. Keywords - Faster R-CNN, Convolution Neural Network, SSD (300x300)