Finding life on the seafloor with AI

Michael Stanley and Harsh Panchal are data science researchers at Lynker Analytics and Masters students at Victoria University. Together they will share results from an international NOAA and NVIDIA GPU Hackathon held in December.

The goal of this project was to detect and classify biology from over 100,000 images captured by an Autonomous Underwater Vehicle (AUV) operating at depth off the United States west coast. The team will explain how object detection and image classification models were developed using Self Training and Active Learning techniques.

They will explore the results of the project and provide personal perspectives on this project as graduate students working in industry.

About Lynker Analytics:

Lynker Analytics is a New Zealand company with specialist expertise in data science, geospatial analytics and machine learning. They have developed sophisticated tools to extract new knowledge and insights about the earth's surface and built environment from satellite imagery, aerial photography, drones, Lidar and other data sets.

Using neural networks and related AI techniques, Lynker Analytics can monitor and quantify change in the environment at local, city or national scales to support better decision making.