Qianwen Luo
Pronouns: she/her/hers
Research Mentor(s): Nathan Fox
Research Mentor School/College/Department: Sustainable Future Hub / SNRE
Program: UROP
Session: Session 3 (11:00am – 11:50am)
Authors:
Abstract
Plant pathology is an important problem of research interest. The plant pathology challenge 2020 focuses on training a model to classify foliar diseases of apples given the photos of the apple leaves. In this task, data insufficiency and data imbalance should be addressed. To tackle the above issues, we implement a series of data augmentation operations including illumination, contrast adjustment, flipping, rotation, cropping, and blurring to enrich the dataset. We use ResNet as the model framework and use 5-fold cross-validation to train the model. Experiment results show that the proposed methods can achieve good results.
