An Underwater AR based System for Marine Life Detection and Classification for Divers and Tourists

Research Project

Objective/Contributions:

  • Collecting and annotating an extensive dataset of fish images representing the fish species in the red sea.
  • Building a machine learning model capable of detecting and classifying fish species from a real-time video.
  • Building a proof-of-concept prototype for the AR hardware that is capable of capturing the live video of marine life, running the classification model, and displaying the information on a suitable screen.
  • About two high-quality scientific papers will be published in peer-reviewed highly ranked international journals and major conferences in the field.