Innovative Projects Realized

AI for Extraction of Biomedical Signals from Headphones

In an age of so many new wearable devices, e.g., smartwatches, glasses, rings, clothing, and so on, headphones can be recognized as the first widely adopted wearable device. They have been around for more than a century and have been used mostly as an output device for listening to music, or, in the
recent decades, talking on the phone. Even more recently, Ohmic has developed a technology formed by a suite of hardware and software solutions that enables headphones to go beyond their initial purpose. Ohmic has created a dongle to which any wired headphone can be connected to and that transforms any simple headphone into a smart version of itself, enabling gesture recognition (tapping and sliding), user identification, and biometric monitoring (heartbeat). It does that by using a circuit to cancel the incoming audio signal while amplifying and processing the signals of interest.

While the prototype can already read biometric signals, user activities like walking and running interfere with the desired signal causing a reduction in the signal-to-noise ratio (SNR). These so-called motion artifacts are a very well-known problem1 for wearable devices in general, however it can be mitigated by use of artificial intelligence to extract desired features, e.g., heart rate and heart rate variability. The latter being used to detect human emotions.

In the current context of making any headphone smarter, the intern student will be researching and developing the most appropriate machine learning strategies that can be used to extract biometric features, more specifically to ensure signal integrity, to understand the needs for future development,
and to lay the basis for the path from R&D prototypes to mass market products.

More precisely, the main tasks will be :
? Working on a literature review and comparison table of machine learning techniques for
biometric signals, highlighting their respective advantages and drawbacks.
? Developing AI algorithms for feature extraction of biometric signals, such as heart rate, heart
rate variability, respiration rate, and in-ear canal characterization.
This project is stand-alone and no other projects are planned at this time.

Permeability-based leakage model to describe functional failure in composite storage vessels for alternative fuels

Pressure vessels made entirely from fiber-reinforced polymer can generate substantial cost and performance improvements leading to their greater acceptance as a storage medium for pressurized alternative fuels (e.g. hydrogen). Inherent anisotropy and inhomogeneity of fiberreinforced composites usually induces functional failure (i.e. leakage) that precedes structural failure by bursting. Functional failure is attributed mainly to transverse matrix micro-cracking. The complex material behaviour and diverse failure mechanisms associated with composite pressure vessels cannot appropriately be modeled by existing failure criteria. The goal of the proposed project is to develop permeability-based analytical models to evaluate the performance of composite vessels under monotonic and certain fatigue loading conditions.

Conditions environnementales associées à l’habitat du frêne noir en plaine de débordement au lac Duparquet

Le projet vise à évaluer les conditions permettant au frêne noir, une espèce protégée, de se maintenir sur les plaines de débordement du Lac Duparquet en Abitibi dans un contexte où les changements climatiques pourraient entrainer une modification plus ample du régime des crues. Pour ce faire, le projet vise trois sous-objectifs soit i) localiser et géoréférencer les peuplements de frêne du lac Duparquet (dimensions), ii) géoréférencer les transitions écologiques (zone de végétation) survenant à l’intérieur des peuplements et iii) déterminer à l’intérieur de ces zones la sévérité des crues enregistrées par les arbres et ce par l’analyse des cernes de crues. L’analyse des cernes de croissance permettra d’établir dans chaque zone l’intensité/durée des crues. Après avoir construit une base de données géoréférencées et en recoupant cette information avec la cartographie LIDAR, les résultats serviront à évaluer comment les changements dans le régime des crues pourraient influencer la pérennité de ces peuplements exceptionnels situés près de leur limite nordique de distribution.

Automatic Seizure Detection, Prediction, and Mitigation Using Minimally Invasive Implantable Bioelectrical Sensors

Epilepsy affects 65 million people globally, with 30% unable to control their seizures with medication.
Implantable devices that control brain activity via electrical stimulation are an alternative treatment option;
however, existing FDA-approved devices only render 13% of patients seizure-free for up to a year.
NerveX Neurotechnologies has developed ANIMO (Adaptive Neural Implant for Modulation and Observation), which solves the key problems causing limited efficacy in existing solutions. By leveraging machine learning insights, ANIMO can offer valuable insights in predicting when a seizure may occur. The present project will be concerned with improving seizure detection algorithms and customizing ANIMO to individual patient needs.

Speech enhancement and recognition with generative adversarial network

While taking foreign language tests, people may record responses with different environments and equipment. sometimes the recording may not be very clear. The low-quality audios can lead to unusual results in speech recognition and scoring by the scoring systems. Audios with a higher resolution (sample rates) contain richer information since greater frequency ranges can be represented in the data, capturing greater level of detail and texture to produce high-quality audio, such as sibilants and fricatives. Pearson would like to improve the quality of existing data by transforming audios with lower resolution in the dataset to audios with higher resolution, in order to better train and develop a more robust system for the automated speech recognition machine. Improving these systems would greatly benefit Pearson’s competitiveness in the market and would also contribute towards expanding the boundaries of knowledge in speech enhancement.

Automatic alignment of subtitle and audio track for TV broadcasting

Subtitles are typically generated postproduction and might be mis-aligned with the audio track. Alignment of subtitles to audio is a task which aims to produce accurate timings of the words within a subtitle track given a corresponding audio track. The alignment task is of significant relevance to Bell media. Semi–automated solutions exist, and they involve different degrees of human involvement in post-edit. Automatic subtitle alignment will reduce production time and human labor costs. This research project will look at automating the subtitle alignment.

Effective experimental and modelling approaches to investigate CO2 capture with IL/water

The increasing concentration of CO2 has caused various adverse environmental effects on the Earth’s oceans, land, and atmosphere, leading to a worldwide consensus on the necessity of action and commitment to emission reduction. Carbon Capture and Storage (CCS), which involves capturing CO2 resulting from industrial processes for storage in underground geological formations, has proven to be an effective method for CO2 emission mitigation. CO2 absorption using chemical solvents such as amines is a prevailing but technically challenging and rather low-efficiency method of carbon capture. Ionic Liquids (ILs) provide a green alternative for chemical CO2 capturing. However, their typically high viscosity and uncertain selectivity pose challenges to their application. This research is intended to improve the techno-economic viability of carbon capture through CO2 absorption with ILs. We integrate the process design in customizing task-specific ILs through a combination of experimental, mathematical, molecular modelling, and machine learning approaches.

Encrypted IoT Network Traffic Analysis for Cybersecurity

Use of clear-text for network communication is quickly becoming obsolete. End-to-end encryption of mobile
applications and proliferation of TLS based encrypted traffic are fueling the growth of encrypted traffic on the
Internet. Many applications tunnel traffic over HTTPS. Today, over 80% of Internet traffic is encrypted. With the
recent advancement of TLS handshake, traditional mechanisms such as Deep Packet Inspection (DPI) and TLS
fingerprinting are unable to inspect traffic in the traditional manner. As a result, operators are unable to block
illegal applications, enforce policies, and detect cyber threats hiding in an encrypted traffic. In addition, extensive
deployment of hyperconnected devices such as Internet of Things (IoT) has created a new attack surface. Further,
an emerging generation of IoT devices will support encryption, making detection of rogue IoT devices a challenge.
In this project our goal is to design, develop and evaluate a machine learning based platform for encrypted network
traffic analysis in order to detect rogue IoT devices and malicious behaviours.

A Microfluidic Biosensor for Monitoring Post-Operative Drainage Fluid

Anastomotic Leakage is a common complication that can occur shortly after surgery and lead to increased mortality rates if it is detected too late. The current method of detection relies on monitoring clinical symptoms and visual inspection of the surgical drain fluid, but this process is often delayed and reactive. NERv Technology Inc. is dedicated to developing a system that can continuously monitor the contents of the drain output for a long period of time (i.e. ~ 10 days), and provide early detection of key biomarkers that indicate anastomotic leakage. Amylase is one of the key biomarkers associated with anastomotic leakage. This project proposes to develop a microfluidic-based amylase biosensor that is compatible with NERv’s existing system for inline monitoring the surgical drain output by leveraging Dr. Ren’s expertise and technical know-how in microfluidics and electrical sensing and NERv’s expertise and knowledge about anastomotic leakage problem.

Development of a plant-based production system

Rare cannabinoids are currently gaining popularity for their potential uses in pharmaceuticals, therapeutics and cosmetics. Traditional cannabis companies are currently unable to produce these compounds and the demand is expected to increase significantly in the future. This highlights the need for new, innovative solutions to solve this problem. Using synthetic biology, the intern will develop a prototype for the company’s plant-based production system that will be used for the production of rare cannabinoids. This results of this project will be used to file the provisional patent application and blaze the trail towards implementation and commercial application through discussions with established companies. These complementary companies will be able to partner with TerraPlant or license (non-exclusive licenses) its technology.

An Examination of Mental Health Leadership and Social Advocacy by Post-Secondary Students with Lived Experience in Atlantic Canada

The crisis in university student mental health has intensified with COVID-19. Support for students with mental health and illness experiences and identities is now more important than ever. Prior to the pandemic students were taking matters into their own hands and working together to advocate for improved learning environments and educational experiences. This project will examine how participation in campus-based mental health advocacy can support more inclusive and equitable experiences of university education in Atlantic Canada. Results will raise awareness of student-led mental health advocacy in universities in Atlantic Canada, mobilize current evidence on university student led mental health advocacy, and identify influences that support positive experiences and outcomes from the perspectives of student advocates themselves.

Évaluation de l’ampleur économique de la « véloconomie » au Québec

L’objectif de la recherche est de quantifier l’importance économique de la filière cycle, comprise dans son sens large (d’où le néologisme de « véloconomie »). Le Québec dispose de précieux atouts en la matière sur son territoire avec une demi-douzaine de fabricants de vélos et la plus importante véloroute en Amérique du Nord. Cette tendance va s’accentuant, puisque les entreprises en cyclologistique se multiplient depuis les cinq dernières années, les services de vélos en libre-service se développent et les magasins de vente et de réparation de vélo connaissent une prospérité record.
Outre le poids économique « direct » de l’industrie du vélo, tel que mesuré par le chiffre d’affaires et la création d’emplois, de nombreux aspects économiques liés au vélo ne sont pas immédiatement chiffrables, comme l’amélioration de la santé (mentale et physique) des usagers, la réduction de la pollution et l’augmentation du tourisme, qui nécessiteront des techniques spécifiques d’évaluation.