In order to better understand the role of Canada’s vast forested area in our country’s carbon budget, further work is needed to monitor ‘hot spots’ of carbon activity – the boundaries between land and lake, and how these landscape positions will react to a changing climate. Further, less is known about the timing around the activation of these hot spots outside of regularly spaced traditional monitoring practices.
Graphics Processing Units (GPUs) are usually employed to quickly render images on everyday computer screens, and do so quickly and efficiently for relatively little cost compared to the use of Central Processing Units (CPUs) that make up the “brain” of the computer. Modern GPUs are able to do hundreds or thousands of simultaneous calculations; rewriting conventional computer problems in the language of GPUs offers the po-tential to dramatically decrease the computation time for complex problems such as Electromagnetic Trans-mission (EMT) simulations.
The Canadian Arctic is warming at an alarming rate. The coastal community of Pangnirtung, Nunavut has long witnessed and experienced the reality of climate change. Country food is the main resource for Panniqtuumiut all year round and practices related to hunting and fishing are key to family and community well-being. Local organizations and community members contribute to numerous academic studies and endeavours devoted to climate change.
The problem considered in this work is how to produce highly accurate and consistent land-use/land-cover (LULC) maps significantly faster than current semi?automated methods for use by Manitoba Hydro. The goal is to improve the ability to produce maps quickly and efficiently as priority needs arise. This project will use an approach for automated LULC mapping from satellite images using deep learning methods pioneered by the applicants. By classifying each pixel in a satellite image into LULC categories using neural networks, rapid and accurate LULC maps can be successfully produced.
We envision a future where it will be possible to lavish the same attention on individual plants in a large prairie crop farm as one might on those in a backyard garden. As camera sensors shrink in size, and self-driving vehicles continue to improve, such an idea is no longer the realm of science fiction. The remaining piece of the puzzle, however, is the need for a very large number of pre-identified images of crop plants and weeds with which to train a computer to recognize one from the other.
Graphics Processing Units (GPUs) are usually employed to quickly render images on everyday computer screens, and do so quickly and efficiently for relatively little cost. Modern GPUs are able to do hundreds or thousands of simultaneous calculations; rewriting conventional computer problems in the language of GPUs offers the potential to dramatically decrease the computing time for complex problems such as Electromagnetic Transmission (EMT) simulations.
Funded through a series of NSERC partnership grants, Cubresa Inc. have been working with the applicants to develop technologies for pre-clinical imaging for positron emission tomography (PET) imaging and PET friendly radiofrequency (RF) coils for magnetic resonance imaging (MRI) of small animals.
Grasslands are one of the most endangered habitat in North America. In Manitoba, over 90% has been lost in the last 100 years and with it a suite of prairie adapted species. The Poweshiek skipperling is one such species which in recent years has plummeted in abundance for unknown reasons. Less than 500 individuals remain in the wild and the grasslands of southeastern Manitoba represent one of the species last strongholds.
2012 marked a pivotal milestone in the field of neural networks. The intersection of general purpose computing using Graphics Processing Units (GPUs), labelled big datasets, and very large neural networks (called deep neural networks) enabled a break-through in machine learning that has led to impressive results in many fields and applications, such as self-driving vehicles and real-time language translation. Recently, the advances offered by these techniques have been applied to the areas of music and speech synthesis, which have opened up exciting new areas of applications.