Transfer learning for semantic segmentation of fungal growth images to accelerate high-throughput fungicide development

There is a growing need to develop new crop protection products that are more effective against fungal disease and mitigate pathogen resistance. The fungicide product development process is time consuming and expensive and there is a great opportunity to reduce costs and time by using machine learning. In this project, an existing deep learning algorithm for quantifying fungal growth in 96-well plate images will be transferred to new pathogens to extend Terramera’s automated high-throughput screening pipeline. This application will eliminate time-consuming manual analysis and accelerate the in-vitro testing stage in Terramera’s product development.

Faculty Supervisor:

Martin Ester

Student:

Partner:

Terramera Inc

Discipline:

Computer science

Sector:

Agriculture; Manufacturing; Professional, scientific and technical services

University:

Simon Fraser University

Program: