Sullivan R. Steele

The goal of this project is to assess the feasibility of using computer simulation to train a machine learning (ML) model for in-tank fish detection in a recirculating aquaculture system (RAS) environment. The project will involve creating simulated underwater images of fish in a RAS environment using advanced 3D modeling techniques. The trained ML model will then be compared against a traditional fish detection model in terms of accuracy and performance. The outcome of this project will provide valuable insights into the potential of using simulated images for ML model training in the aquaculture industry. 

I developed dashboards for technicians to track and interact with automatically captured data easily. Along with this, I also implemented a local database application to generate and organize images and biodata from the project for easy access across the work site.

The Fitbit bridges project is a program presented by WVU in which college undergrads/grads are given a Fitbit health tracker watch and assigned to research into visualization, analysis, and insights using the data available via Fitbit’s website and API. All students in this program are to collaborate and work together to do this research and prepare a presentation as well as a written report on the subject of interest.