Projets novateurs réalisés

The impact of a novel plant-based protein blend on aminoacidemia and muscle protein synthesis

There is no plant-based protein (PBP) that compares to whey protein isolate (WPI) on the growth and maintenance of muscle mass, as WPI contains all the building blocks of protein, known as amino acids (AA). The most important AA when it comes to muscle growth is leucine, which is not as high in PBPs. In order to close this gap, we will be comparing a new PBP blend with sufficient leucine to WPI in 2 studies to confirm the new PBP blend’s safety and tolerability and its impact on muscle growth. Results from these studies will be used to advance knowledge on the impact of PBP blends on muscle health while guiding the development and application of PBP blends in Canada.

High-fidelity, Directable Animation Transfer for Dubbing

The entertainment industry delivers media products around the world to people speaking a large number of distinct languages. To accommodate different languages, the producers typically re-generate the sound stream in different languages in a post-production step, but they leave the same video footage resulting in the desynchronization between the audio and video content, creating an unnatural effect. In this project we are developing and extending a set of tools and techniques to naturally synthesize the video footage to match the audio track eliminating the visual artifacts. Such methods and tools will help create higher quality video and audio content that can be easier “translated” to any language increasing the competitive edge of the company as well as the dubbing industry as a whole.

Multi modal information extraction for visually rich documents

The main objective of this project is to extract information from documents (Name, issue date, address, account number, etc.) that have diverse set of layouts (bank statements, utility bills, credit card statements, etc.). This problem is challenging because of diversity in layouts, variety of languages and complexity of template structures, which makes traditional NLP approaches difficult to use. The goal is to develop a solution that can make use of not only textual but visual layout information to extract information. Intern will be involved in reading research paper pertaining the problem, analysing image datasets and write code to run experiments on public and proprietary datasets to improve the accuracy of the system. This project is part of Jumio’s Document Verification product, that provides companies with services that verifies customer information through the analysis of their documents.

INO : Détection automatique de menaces dans des images issues d’une caméra térahertz (THz).

INO est en train de développer un dispositif «?see-through?» permettant d’inspecter sans contact une zone prédéfinie sur une personne. Ce genre de dispositif serait déployé aux endroits où un contrôle des personnes est requis afin de déterminer s’ils transportent des articles menaçants sous leurs vêtements. Citons à titre d’exemple les aéroports. L’utilité de ce dispositif est de résoudre différents enjeux reliés à la fouille corporelle (parties intimes, signes religieux, risques sanitaires, etc.).
Dans le but d’accélérer et rendre plus efficace la procédure de contrôle, INO veut explorer les possibilités d’automatiser la détection de menaces. La portée du projet de stage est donc de développer une ou plusieurs approches basées sur l’intelligence artificielle qui permettront de classifier les données issues des scans en menace (couteau, fusil, etc.) ou non-menace (monnaie, montre, collier, etc.).

Histoire des paléo-îles du lac proglaciaire Ojibway (partie 2)

L’Amérique du Nord était recouverte de glace durant le maximum glaciaire il y a de cela 18 000 ans. Par la suite, le climat s’est réchauffé et les glaciers se sont progressivement retirés. Ce retrait des glaciers a été associé au développement d’un gigantesque lac couvrant la région actuelle de l’Abitibi-Témiscamingue. Ce lac dénommé Ojibway, était peu profond et présentait des îles, devenues aujourd’hui des collines comme les collines Abijevis du Parc National d’Aiguebelle. Les travaux de recherche entrepris en collaboration avec le Parc visent à reconstituer le paysage de ces îles à l’époque où le reste de la région était recouvert par le lac Ojibway. Les résultats permettront de mieux caractériser l’évolution de l’environnement de la région et d’alimenter les stratégies de gestion et de conservation des massifs forestiers du Parc national d’Aiguebelle.

Parking Occupancy Inference With LiDAR Sensors

Parking is a cumbersome part of auto travel in urban areas, primarily due to lack of information on the location of available spots. Sensors can be deployed to detect occupancy, but they often fail due to their high costs and detection inefficiencies in outdoor spaces. This project pursues a feasibility study of using LIDAR sensors, which overcome some deficiencies, for parking detection. LIDARs have a wide field of view, are robust to outdoor disturbances, and can be provided at cost given the recent advancements in the autonomous vehicle industry. The steps of the project include: (i) review types of LIDARS and their properties, (ii) process data collected from LIDARS, and (iii) develop data analytics solutions for interpreting LIDAR data.

Validation d’une plateforme numérique intégrée de sensibilisation Grand Public à la cyber-sécurité

Le présent projet vise le développement de modules de formation et d’éducation novateurs ayant pour objectifs d’améliorer les habitudes cyber-hygiéniques. Le but premier de ces modules de formation est de développer un sens de cyber-responsabilité, ainsi que des habitudes numériques sécuritaires. Ces modules de formation seront ludiques, conviviales et adaptées à la réalité des usagers afin de soutenir l’engagement des utilisateurs durant le visionnement et ainsi favoriser une meilleure rétention du message à travers le temps. L’objectif étant que les utilisateurs retiennent le message et utilisent les connaissances et les compétences acquises en comportement observable dans la vie de tous les jours afin de diminuer les risques de fraudes et réduire les risques financiers.

Minimizing the Impacts and Maximizing the Benefits of Marine Shipping Activities for Arctic Communities Through the Use of Traditional Knowledge

Interest in Arctic shipping is growing as sea ice is melting and as other industries are developing. An increase in shipping presents risks to both the natural environment (from the movement of ships and the products that they carry) and to the local communities that rely on those healthy environments for food and other products. However, the local Inuit people in Nunavut also hold a certain body of knowledge referred to as Inuit Qaujimajatuqangit (IQ). This knowledge includes critical information about the land and water, its use, and important cultural practices. IQ must be considered in development projects in the Arctic in order to ensure that any possible negative impacts from the project can be minimized or avoided. The identified issue is that managers do not know how to effectively incorporate IQ into the decision-making process.

Anonymisation et désensibilisation de données

L’enjeu de protection de la confidentialité des données personnelles sensibles et de leur utilisation par le secteur privé pose de plus en plus de défis. L’adoption par le gouvernement du Québec de la loi n°64, pour encadrer les dispositions législatives en matière de protection des renseignements personnels, remet la question au cœur des débats. Le principal défi technologique devient celui de l’exploitation de la donnée individuelle au bénéfice du collectif sans sacrifier sa confidentialité : Comment tirer profit des données clients sans menacer leur vie privée ?

L’anonymisation des données est, à ce titre, une stratégie courante, pour récolter les avantages de l’exploitation des données pour les individus et la société en général, tout en atténuant les risques pour les personnes concernées.

Ce projet propose ainsi d’explorer les méthodes d’anonymisation existantes pour répondre au double enjeu d’exploitation et de préservation de la confidentialité des données personnelles.

Modeling Ekona’s H2 production and carbon capture process: Phase 2

Ekona Power Inc is developing a technology to convert natural gas into H2, electricity, and a solid carbon stream. The solid carbon stream will then be sequestered. Their approach involves two reactors. The first reactor combusts natural gas and oxygen to provide the process heat to the second reactor. The second reactor heats natural gas to convert the natural gas via pyrolysis to hydrogen and solid carbon. This proposal focuses on the second reactor.
The proposed modeling approach will allow Ekona to better understand its pyrolysis design and anticipate the performance of the commercial unit. The model can be used to identify process changes that will improve process performance and identify key tradeoffs.

Preserving Affordable Rental Housing: Tenant Opportunity to Purchase, Housing Cooperatives, and Land Trusts

Housing is becoming increasingly unaffordable in Canadian cities and tenants are struggling. Financialized landlords and other investment companies are depleting the affordable rental stock in the private market much faster than new social rental units are being created. One solution is for tenants to have the opportunity to collectively purchase their own buildings when landlords decide to sell. Tenants could convert buildings to limited-equity housing cooperatives leased from a land trust for the long term. In this way, tenants avoid displacement and achieve housing security, and buildings are converted to permanently affordable housing for future generations. Such programs already exists in several American cities, under Tenant Opportunity to Purchase (TOP) policies. This research, conducted in partnership with the Hamilton Community Land Trust, will evaluate the feasibility of such a program in Hamilton, Ontario. If successful in Hamilton, similar programs could be adopted by land trusts across Canada.

Designing for better novice assembly: towards energy efficient novice building

This research is about technologies for ‘DIY’ or ‘novice’ home-construction. The research asks if high energy efficiency homes could be built by novice and non-expert builders. Energy efficient construction is an increasingly expert craft: with specialized knowledge, tools, and practices now required to achieve excellent building energy performance, this presents barriers to building efficiency for novice builders. This research will identify the barriers and opportunities for evolving more energy efficient novice building processes by working with Habitat for Humanity (an organization that specifically leverages novice labor) and Passive House 43 (a Toronto based energy efficient home designer) to understand Habitat’s current construction practices, and how their use of novice labour shapes energy efficiency in the completed homes. Using human centered design and analysis methodologies (such as stakeholder interviews, process blueprinting, and builder/resident journey mapping), researchers will work alongside the design and construction team for a live Habitat for Humanity project build to envision feasible possibilities for more energy efficient novice processes. The results of this research will form a set of design and process recommendations that can be used on future Habitat for Humanity projects to address the need for affordable energy efficient home-building.