VitisGen2 researchers describe improved deep learning method for quantification of Grape PM at ASABE Meeting

Deep learning-based saliency maps for the quantification of grape powdery mildew at the microscopic level
VitisGen2 team members Tian Qiu, Anna Underhill, Surya Datta Sapkota, Lance Cadle-Davidson, and Yu Jiang recently presented a paper at the 2021 American Society of Agricultural and Biological Engineers (ASABE) International Virtual Meeting that urges the breeding of Powdery Mildew (PM) resistant crops and how to overcome the bottleneck in image analysis at the microscopic level.  The researchers described a new method developed using high-throughput phenotyping to more rapidly quantify disease. The method uses the Blackbird phenotyping robot to automate and improve the speed of processing samples, and could lower costs.