A New Dataset of Millions of Labelled Images for Machine Learning Applications in Digital Agriculture


Thursday, January 28, 2021

Online Webinar & Q&A
Livestream Link – https://www.youtube.com/watch?v=aSfIEC9ZzNQ

Q&A – Sli.do event code #E974

Dr. Christopher Henry, University of Winnipeg

Abstract: Automatic plant classification is a required component to enabling digital agricultural applications such as the targeted destruction of weeds by autonomous vehicles. One such scenario might see autonomous vehicles fitted with imaging sensors that operate day and night, inspecting hundreds of hectares of farmland plant-by-plant. The development of autonomous vehicles is well underway and their eventual arrival seems to be a certainty.  However, current deep neural networks that underlay automatic image classification systems still depend on vast quantities of labelled data in order to train and achieve high-accuracy plant classification. This presentation discusses our efforts and progress toward automatically generating and labeling very large numbers of images of crop plants and common weeds for use by academia and industry in their push toward Agriculture 4.0.

Bio: Dr. Christopher Henry is an Associate Professor in the Dept. of Applied Computer Science at the University of Winnipeg and is an Adjunct Professor in the Dept. of Electrical and Computer Engineering at the University of Manitoba. He has held NSERC Discovery funding throughout his career. The focus of Dr. Henry’s work is the theory and application of machine learning, such as his work in developing machine learning data sets for digital agricultural applications. This has led to machine learning collaborations with Sightline Innovation, Northstar Robotics, and DecisionWorks.

Dr. Henry has also pioneered approaches to classify pixels obtained from satellite images using deep neural networks developed for semantic segmentation for the creation of land-use/land-classification maps. The success of this work has led to industry research contracts and grants with GeoManitoba and Manitoba Hydro, as well as collaborations with researchers at the Norwegian University of Science and Technology. This work has generated a start-up company, DeepGeo, to commercialize this work. DeepGeo was created through the Creative Destructive Laboratory in the Rotman School of Management at the University of Toronto and secured its first contract with Manitoba Agricultural Services Corp on a project to perform crop yield prediction from satellite data.

Dr. Henry has also been collaborating for many years on GPU-based computing initiatives, including collaborations with Manitoba Hydro, Ubisoft, and Manitoba Hydro International.  He has worked to establish a provincial consortium of researchers in high-performance computing that ran several roundtable meetings with industry as well as two conferences; he has co-founded the Applied Parallel Computing and Collaborative Research Laboratory, and co-led the process to establish the University of Winnipeg as an NVIDIA GPU Education Centre.