Projets novateurs réalisés

Calibrated and automated recognition of barcodes in complex, highly variable scenes

The main objective of this project is to investigate, develop and evaluate state-of-the-art artificial-intelligence (AI) algorithms for the automatic recognition of extremely diversified barcode instances, in complex, highly variable and non-controlled scenes. The application area is of high interest to the industrial partner, and the ultimate goal is to build a state-of-the-art, automated and well-calibrated barcode analysis system, which meets challenging industrial expectations of high accuracy and computational efficiency, while providing measures of confidence of the predictions (to mitigate the risks associated with potential errors). The recent years have witnessed substantial technical advances in deep learning (e.g., convolutional neural networks), which promises to address such complex image analysis problems. The project will involve one intern (a PhD student), whose objective is to adapt, extend and develop state-of-the-art convolutional neural networks for training automatic recognition models, while accounting for specific industrial challenges: high-accuracy requirements, the need for effective uncertainty estimation to mitigate the high cost of errors, and the need for active learning and exploring the diversity of large-scale unlabeled data. These model-calibration, semi-supervision and active-learning problems are of great interest to the industrial partner and to the computer vision community in general.

Examination of Foundations and Disbursement Quota in Canada

The purpose of this project is to provide a study of charitable spending by members of Philanthropic Foundations Canada, document the historical impacts brought about by changes to Canada’s disbursement quota (DQ) policy, and recommend changes to the T3010 form. The intern will lead knowledge generation in the form of member engagement, issue briefs, and attending conferences. Throughout this process, the intern will encourage Canadian charitable foundations to embrace DQ-related policy positions that promote sustainability and the success of the charitable sector as well as supporting the application of best-practices. This research will provide a foundation for further study.

Understanding minke whale migration, feeding, and population structure in the North Atlantic

Minke whales are common in the North Atlantic, but there are huge gaps in scientific knowledge about the species. Through collaboration with Mériscope, we will attach satellite tags to minke whales that feed in the St Lawrence Estuary (SLE) in the summer to track where they migrate throughout the rest of the year. At the same time, we will collect tissue samples from the SLE minkes, which we will analyze in comparison with minke whales across other parts of the Atlantic. By looking at their genetics, we can identify differences and connectivity between groups, and using other chemical signals in their tissues, we can shed light on the whales’ diet, which in turn can provide information about the larger ecosystem. Overall, having a better understanding of minke whale biology will be essential both to their conservation and to understanding how climate change is affecting the North Atlantic marine landscape.

Improving power production by better recognizing power quality issue events with machine learning

Solar energy offers a green energy future both for Canada and the world. To best collect this energy, solar farms, collections of solar panels, are often distributed to ensure efficient local collection. Weather is often a challenge for production; however, the failure of components can also adversely affect this as well. These failures can be difficult to track and predict; in this project, we propose to develop tools to help operators expect events that could lead to power losses and improve solar energy harvesting using both standard analytical tools and machine learning. This will lead to improved economic benefits for Canada and production efficiency improvements.

The effects of stress during white sturgeon early life history on larval physiology, development and olfactory sensitivity

Elevated levels of stress hormones (i.e. cortisol) provide a potential mechanism by which substrate condition affects larval development and survival in the endangered white sturgeon. The possibility that stress hormones mediate larval response to substrate conditions will be investigated by artificially elevating cortisol levels and identifying effects on larval physiology. Cortisol levels of wild caught adult fish will also be investigated to determine the effects of capture and handling on spawning success and larval quality. Two final projects will investigate whether substrate condition (habitat stress) affects larval olfactory development and/or thyroid hormone levels. Both of these factors may be linked to the mechanism of imprinting by which fish return to their birthplace to spawn (e.g. salmon). A better understanding of factors affecting larval survival and quality will help both project partners implement both hatchery and habitat based components of the species recovery plan.

Functional and structural characterization of clinically relevant amyloid fibrils

Irreversible, progressive, and incurable neurodegenerative diseases, such as Alzheimer’s disease (AD) or Parkinson’s disease (PD), constitute severe challenges for the modern and, in particular, for the healthcare system. Recent reports suggest that worldwide >50 Mio. people and >10 Mio. people are living with AD or PD, respectively. Both diseases are related to protein misfolding and aggregation towards elongated and helical amyloid fibrils. However, although amyloid fibrils are considered pathological hallmarks for AD and PD, the mechanisms of how such fibrils hamper normal cellular function are largely unknown.

The long-term objective is to shed light on the molecular basis of neurodegenerative diseases, in particular AD and PD. Therefore, we integratively combine high-resolution structure determination of clinically relevant amyloid fibrils with modern biochemical in vitro and in vivo characterization. Our approach may elevate the understanding of the pathologic mechanisms of neurodegenerative diseases, which is essential to promote the rational search for novel, innovative drugs.

Cross-frequency coupling in organoids

By growing stem cells in the right conditions, we are able to create miniature brain tissue structures called organoids. These organoids have many properties that make them similar to adult brains, which makes them a valuable tool to study how brain tissues might be affected by drugs, or how brain diseases like epilepsy can occur. One method that doctors currently use to choose which drugs to give to an epileptic patient is to examine the electrical activity of an epileptic brain for the kinds of rhythms that are present and to test different anticonvulsants over time to find one that works. Our project aims to measure the electrical activity of organoids for the same kinds of rhythms that might be present in an epileptic brain, to see if these same rhythms can exist in organoids. In order to make sure we mimic what happens in epilepsy, we can add a chemical convulsant to stimulate epilepsy-like behavior from the organoids.

In vitro effects of high irradiance delivered by laser and polywave light-curing units on the simulated pulp temperature rise and mechanical properties of bulk-fill resin composites

Dental light-curing units (LCUs) are used every day in most dental offices. If this LCU is misused, this dental work can fail prematurely causing pain, post-operative sensitivity, pulpal death, and economic hardship. In Canada, dental curing lights are Class 1 medical devices, and Health Canada does not test dental curing lights before they are used on patients. However, LCUs can easily cause soft tissue burns or retinal damage if used incorrectly. In the past 15 years, dental LCUs have become more and more powerful and offer shorter and shorter curing times (previous curing times were 60 s, now they are as short as 1 s), but the success of such short exposure times is questionable.
There is no information published on this recently introduced Monet laser LCU, only manufacturer’s claims. This project will evaluate if the light emitted from the Monet laser-based LCU can effectively cure plastic-like restorative materials (resin composites) without causing any risk due to the heat. Extracted molars with deep cavities will be connected to a device that simulates pulpal fluid flow inside the tooth.

IC prototype for decorative fingernail-attachable flexible e-ink display communicating using NFC

AIM Colours is looking to revolutionize self-expression by combining cosmetics with technology through the creation of eCosmetics. We envision eCosmetic applications as devices that can be continuously colour-manipulated from any mobile device, creating a world where anyone can be empowered by make-up instantaneously. This project requires the seamless integration of engineering and science to ensure the development of the decorative wearable electronics field. The purpose of this project is to develop the physical prototype for the ePolish. ePolish are technology-enabled press-on nails that allow users to interact with the appearance of their nails directly from their smartphones. This revolutionary version of nail polish will allow users to save time and avoid chronic exposure to chemical solvents.

Multiscale Bat Habitat Selection and population trends at Maternity Colonies and Hibernacula in Southwestern Quebec

Bats are fascinating animals that provide many ecological services by regulating insect populations. They can eat about 600 insect per hour! Since the introduction of white-nose syndrome in 2006, bat populations in North America have become increasingly threatened. This fungal disease has caused dramatic decreases in many bat populations. Without a cure for white-nose syndrome, the protection of bats and their habitat might be the best way we can help them. However, since bats are small and nocturnal, their preferred habitats are understudied. In partnership with the Kenauk Institute, we will create a portrait of bats habitat selection at summer roosts and at winter hibernacula in Quebec. We will use citizen science data to collate information on bat population trends and map known maternity roosts and hibernacula to study the landscape surrounding them. The result will help us understand where and which species of bats live on the Kenauk Reserve and what habitat we should protect to help them. It will help land managers like the Kenauk Institute to take informed decision when managing land with known bat presence.

Neuroinflammation in Huntington’s Disease

Huntington’s disease (HD) is a genetic neurodegenerative disease characterized by cognitive, motor, and psychiatric disturbance ending in death 15 to 20 years after the onset of the first symptoms. The genetic predictability of HD makes it potentially the most tractable of the neurodegenerative diseases for early intervention, with the optimum time for the introduction of disease-modifying therapies being before the onset of symptoms. Despite discovering the underlying genetic mutation more than 20 years ago, treatment remains focused on symptomatic management. Given the devastating emotional costs of this disorder and the significant financial burden, these individuals impose upon society, developing a treatment regime for HD is a recognized priority. To treat a disease, it is necessary to understand the mechanisms behind the symptoms. So, in this project, we will investigate the role of inflammation in HD neuropathology.

Financial Analysis of Digital Assets: Inter-relationship with traditional assets and improving digit asset modeling over traditional financial assets

In May 2021, the total market capitalization of digital assets such as cryptocurrencies and Non-Fungible Tokens (NFTs) exceeded $2.4 Trillion (USD), tripling in the past 3 years. Yet, digital assets are widely misunderstood even by experienced investors, primarily because the technologies empowering digital assets are new and are rapidly evolving. Hence, scientific approach to understanding digital assets is necessary, given their increasingly significant role in financial markets. The primary goals of this project are (a) to help bridge the gap between the analysis of traditional financial market and the rapidly evolving world of digital assets; (b) to create tangible and intangible measures of value for digital tokens; and (c) to examine and analyze the data transparency provided by blockchain-ledgers that can improve digital asset modeling over traditional order-books.