This project will build a new kind of AI question-answering system where several specialized “agents” work together under an adaptive controller. Instead of a single model answering a question in one shot, Omar’s work designs a coordinator that plans the steps, decides when more information is needed, and measures how confident the system should be […]

This project aims to create a simple and reliable method to turn 3D scan data from modular construction factories into BIM models using unsupervised learning. By focusing on basic geometric patterns in point-cloud data, it reduces the need for manual labeling and helps build a semi-automated Scan-to-BIM workflow. The interns will develop a full pipeline […]

This project focuses on understanding dengue transmission and testing a promising mosquito-based intervention, called Wolbachia, to reduce its spread. The aim is to develop a computer model that simulates the life of individual Aedes aegypti mosquitoes and their role in transmitting dengue. This model will include mosquito behaviour, the effects of Wolbachia on mosquito reproduction, […]

Radiation therapy treatment planning is a critical and highly complex step in cancer treatment, requiring significant expertise and time to adapt and personalize treatments to each patient’s condition. Current planning methods often rely on manual expertise and standardized approaches, which can limit the degree of personalization and contribute to variability in outcomes. This project addresses […]

The Hierarchical Knowledge Distillation and Pruning (HKDP) framework project we propose addresses the growing demand for real-time data processing in cybersecurity applications. Traditional centralized machine learning approaches are hindered by high computational requirements and privacy concerns, but federated learning (FL) offers a solution by enabling decentralized model training, maintaining data privacy, reducing energy consumption, privacy, […]

Notre projet de recherche vise à contribuer au développement d’un outil d’analyse et de prévision des risques médiatiques environnementaux, sociaux et de gouvernance (ESG), et ce, à partir de l’intelligence artificielle. Plus précisément, l’objectif est d’améliorer la prévision d’événements médiatiques néfastes à la réputation des entreprises. Pour ce faire, nous proposons notamment d’utiliser les évaluations […]

The project uses large language models (LLMs) and information-theory measures to quantify how people think in real conversations. The goal is to create a pipeline that quantifies informational content (the purpose of the utterance is to provide/inquire information) and relational content (the purpose is to build a stronger relational bond with the other speaker) in […]

Ce projet de recherche vise à mieux comprendre comment l’intelligence artificielle générative peut être utilisée en éducation de manière réfléchie, responsable et adaptée à différents contextes culturels. Le stagiaire collaborera avec des chercheur·es et des formateur·rices universitaires aux Philippines afin de partager des cadres développés au Canada et d’échanger sur leurs usages possibles en enseignement […]

The research addresses the fundamental trade-off in financial econometrics between computational efficiency and predictive accuracy under varying market conditions. While existing literature has established hybrid models with high predictive capabilities, these often incur high computational costs that may not be necessary during stable market regimes. Conversely, traditional regime-switching models like the MS-GARCH utilize a Markov […]

Mon projet de recherche vise à rendre la justice numérique plus accessible et humaine en adaptant les outils technologiques aux capacités de compréhension des citoyens. Dans le cadre de ce stage au Laboratoire de cyberjustice, j’interviens en tant qu’élève-ingénieure en cognitique pour analyser comment les utilisateurs interagissent avec les systèmes judiciaires automatisés. Mon objectif est […]

L’intelligence artificielle joue un rôle de plus en plus important dans le domaine de la santé, notamment grâce aux grands modèles de langage capables d’analyser de vastes volumes d’information médicale. Toutefois, ces systèmes demeurent souvent difficiles à interpréter, ce qui limite leur utilisation dans des contextes sensibles. Ce projet vise à développer des outils d’intelligence […]

This project studies a promising alternative to current autoregressive (AR) large language models: diffusion-based large language models. It aims to make them generate high-quality text more efficiently while maintaining speed and energy efficiency. The research will focus on understanding how these models work, identifying their root limitations, and developing improved methods that balance text quality, […]