Smart autonomous vehicles have now become a reality while the efforts are ongoing to improve the safety, security, efficiency, and performance. In the fast paced digital world, all devices including the vehicles generate a huge amount of data every second which has to be analyzed, stored, and communicated with other devices to reach the next technology milestone. Vehicle-to-Vehicle (V2V), Vehicle-to-Infrastructure (V2I), and Vehicle-to-Device (V2D) communication protocols enable such communications for the smart connected vehicles.
Object detection and classification for surveillance applications via deep neural networks have attracted a lot of interests in computer vision (CV) communities. Accurate and fast CV algorithms can alleviate intensive manual labour and reduce human errors due to fatigue and distraction. In detection problem, the aim is to determine bounding boxes which contain interested objects and classify the category of the detected object. Thus, the detection problem can be formulated as a regression problem to localize multiple objects within a frame.
Every year in Canada over 1.7 million patients are diagnosed with Ulcerative Colitis (UC), and have to go through colonoscopy procedures multiple times for disease detection and treatment monitoring. Trained clinicians use endoscopy facilities and technologies for colonoscopy procedures and unfortunately, the current error rate in disease detection is up to 20%. This project will build a framework that will analyze colonoscopy video streams in real-time, and offers the outcomes to support clinicians to accurately detect UC and monitor patient’s response to treatments.
OERA use hydroacoustic echosounder surveys to evaluate the impact on marine life of tidal turbines in the Bay of Fundy. OERA use Echoview software to read in the raw sensor data (e.g. voltages) and convert it to a visual representation. Echoview contains some algorithms to detect the bottom of the ocean. However, the Fundy data is very noisy from several sources including air bubbles, “entrained air” pushed below the surface of the water, and irregular surfaces on the bottom of the ocean. In order to analyze the survey data, manual pre-processing is currently required to annotate the data.
Various forms of usability testing can be used to optimize interface design and maximize human-computer interaction principles . A well-integrated, intuitive interface has the capacity to improve human efficiency, mitigate errors or lapses and improve situational awareness. Usability methodologies such as heuristic evaluations and cognitive walkthroughs can be performed at any point in the design process to identify areas for improvement. Additionally, in-field usability testing with users of the technology can further pinpoint inadequacies in software or hardware design.
Edge computing is expected to play a transformative role for future AI applications in 5G networks by bringing cloud-style resource provisioning closer to the devices that have the data. Instead of running resource-intensive AI applications at the end devices, we can consolidate their execution at the edge, which brings many benefits, such as eliminating the redundant task processing, running machine learning (ML) tasks with sizeable data sets and running ML tasks in a spatial context that is shared by many devices.
Estimates of the population density of marine mammals in an area and the change in population over space and time are critical inputs for managing the interactions of human activity and mammal populations. Visual surveys from boats, shore stations, and aircraft have served as the basis for most population estimates currently used by managers. However, these survey methods are generally only performed in good weather conditions and require many trained observers. These factors make visual surveys expensive and reduce the temporal and spatial coverage of population estimates.
Model-based reinforcement learning allows AI systems to learn and use predictive models of their environments to plan ahead, achieving tasks more efficiently. The proposed project aims to (i) develop methods for identifying when an uncertain and/or flawed model can be relied on to make plans, and when it cannot, and (ii) implement a method which allows an AI system to explore its environment exactly when exploration will be most useful for improving its model-based predictions and plans.
Real estate information systems (REIS) provide real estate market participants with information that helps inform their decisions. Most prior REIS research has focused on price estimating and forecasting. Although price is important, it is not the only variable of interest.
This research seeks to investigate the needs of REIS users. Who are they? What information do they need? How should that information be presented to maximize its accessibility? How do users’ stated preferences for information differ from their revealed preferences?
Virtial-Gym motion-tracking-based system to exercise regularly and safely at home, guided by the expertise of physical therapists who can remotely monitor their clients’ progress. The innovative exercise grammar that therapists can use to describe personalised exercise regimens for their clients.