Advanced AI for Demand Forecast in Fashion and Apparel Retailing

The ultimate objective of the project is to develop an AI-based framework addressing the forecasting needs of a typical fashion and apparel retailer. The project activities involve development of models predicting demand for particular fashion and apparel items in the context of different customer groups, as well as techniques for identifying fashion trends. The developed methods and algorithms will be able to handle uncertainty, as well as imperfection and missing information. Graph-based formats for representing relevant information will be explored.

Induction Heating Supply

Electrical heating of oil sands allows for production in situ in a mode similar to conventional liquid oil production. This avoids disturbing and redistributing the overburden and production sands. A further advantage of electrical heating is that the power consumption can be matched to the availability of sources such as wind power.

Optimal Design of Switch Reluctance Motor Drives

The main goal of this project is to develop a platform that helps to optimal design of a power converter for switch reluctance motors. The research will focus on multi-domain models and multi-objective optimization routines. Due to the high complexity of developing such a tool, the component models will include loss, thermal, and cost aspects, and the optimization routine will aim to optimize the efficiency and cost of the motor drive design.

Development of Artificial Intelligence Powered Technologies in Computational Pathology to Enable Automated Slide Screening in Whole Slide Imaging - Year two

Advances in Whole Slide Imaging (WSI) and Machine Learning (ML) open new opportunities to create innovative solutions in healthcare and in particular digital pathology to increase efficiencies, reduce cost and most importantly improve patient care. This project envisions the creation of new automated ML tools including the design of a custom Convolution Neural Network (CNN) architecture for whole slide imaging in digital pathology. The custom CNN will be trained to learn different representations of histology tissues so that it can separate healthy from diseased tissues.

Application of time-frequency based techniques to assess Auditory Brainstem Responses in newborn hearing assessment

Automatic detection and classification of the Auditory Brainstem Responses (ABR) is used in newborn hearing screening. Improved detection algorithms will reduce test time, prevent infants with hearing loss from being missed while reducing the number of normal hearing babies referred to diagnostic testing. We have already improved the objectivity of ABR classification in neurological assessments by using Continuous Wavelet Transform (CWT) and Machine Learning (ML). In the proposed project, we seek to validate our findings further to improve the objectivity in the newborn hearing assessment.

Deployment of motion platform control architecture for a high-fidelity driving experience in simulators

Realistic driving experience in motion simulators is a key element for the impressiveness of the VR based simulators. In the driving simulator the free motion of the vehicle is mapped to a motion platform with limited workspace by filtering out the motion and scaling it down with a motion cueing algorithm. These algorithms should be designed in such a way to give a feeling to the users as if they are driving a real vehicle.

Towards an Elastic and Reliable Cloud Resource Management

This work is a holistic automatic methodology for cloud resource management system, which is a corner stone to build any cloud system. Cloud players rely on this to reduce management effort and cloud running cost, by enabling dynamic service access to cloud clients with cheapest price for customer, and high revenue for cloud providers. Customer requirements must be achieved based on their service level agreement like availability of the service.

Topologies and Linearization of High Peak-to-Average Power Amplifiers for Digital Broadcast Radio Applications

Broadcast radio is changing from an analog medium based on frequency modulation (FM) to a full digital broadcast based on orthogonal frequency division multiplexing (OFDM). The high peak-to-average power ratio of the OFDM waveform requires different power amplifier topologies and a high degree of linearity. The research in this project analyzes current amplifier performance for digital radio broadcasting in the FM band, investigates new linearization techniques and explores new amplifier topologies.

Integrity and Analytics of Energy Market Data

ReWatt Power facilitates transactions of green energy derivatives/green attributes such as Alberta Carbon Offsets. The company uses a blockchain technology to manage all transactions on their platform. This way, all parties have access to the transaction history in a database they own, and the transactions are immutable. The first objective of this project is to ensure that source energy measurements are accurate before the resulting green attributes are permanently recorded.

HI-DSR: Hyperspectral Images enhancement via Deep Sparse Representation model

The current non-destructive and fast method of hyperspectral imaging technology are used for different application from remote sensing to medical imaging and food processing. Due to the nature of acquired data which are massive as well as physical consideration depend on the type of application, the careful, fast and accurate data analysis is mandatory to accelerate the usage of HSI technology. To cope with the type of HSI data in which the sparsity assumption is applicable, this project aims to address the HSI data representation though the sparse problem under deep learning approach.