The purpose of this study was to explore the characteristics of developing a machine learning application using synthetic data. The study is structured to develop the application for the purpose of deploying the computer vision model. The findings discuss the realities of attempting to develop a computer vision model for practical purposes and detail the processes, tools, and techniques that were used to meet accuracy requirements.
The sample training images were generated using Leopardi, a synthetic media package developed by John Sutor. 10,000 images of each; Hammerhead, Tiger, Blacktip, and Great White were generated and used for training. Using just over 200 authentic images, the algorithm archived a 92% accuracy in object detection, ideal for shark-busy beaches.
The paper was presented at International Marine, Aviation, Transport, Logistics and Trade, CMATLT001 2021: XV. In, to be held in Amsterdam, Netherlands. Jonathan was awarded Chair of the session and received a Top Paper award for the work.
