Understanding and Improving the Reasoning Capabilities in Vision-Language Models

Vision-language models (VLMs) are considered augmentations of text-only large language models (LLMs) with additional access to the visual world. However, it is known that VLMs typically perform worse than text-only LLMs on reasoning tasks that involve only text (e.g., math word problems), raising the question of whether the benefits of visual augmentation justify the loss […]

Read More
Ventus Analytique Inc. BSI Proposal – Wind Data

Ventus Analytiques Inc. is a Canadian technology company specializing in offshore data systems and analytics, with a focus on marine environmental monitoring and renewable energy. As the company scales its operations and prepares for international deployments, there is a growing need for a cohesive cloud infrastructure and data operations strategy that can support a global […]

Read More
Caswil – Classification automatique des demandes de services en soins à domicile à partir de texte libre

Caswil – Classification automatique des demandes de services en soins à domicile à partir de texte libre (1) Principales activités du partenaire : Caswil est une entreprise spécialisée en services-conseils dans le domaine des technologies de l’information, notamment en migration et transformation de données et développement logiciel. (2) Défis : Caswil accumule des données textuelles […]

Read More
Macro and micronutrient measurements in hydroponics

Environmental conditions (temperature, humidity, light, air movement, CO2 and nutrient levels) are carefully monitored in controlled environment agriculture (CEA) to optimize indoor crop production. Irrigation management is prioritized to ensure nutrient availability to plants, water use efficiency, prevent plant nutrient diseases, maximize crop yield and boost desired phytochemical profiles for a given crop. Automated irrigation […]

Read More
Map Data Analysis for Lawn Service Optimization

This research investigates the fusion of multi-source geospatial data to overcome key scaling limitations in automated property assessment. The primary research problem addresses object occlusion in satellite imagery, where features can obscure ground-level surfaces and prevent accurate area measurement. We propose and evaluate a novel machine learning approach that geometrically fuses satellite imagery with ground-level […]

Read More
AI for Airplane Engine Inspection

This project aims to develop an AI-powered visual inspection system using image data from P&WC’s existing Computer Vision Inspection System (CVIS) to enhance aircraft engine assembly issue detection. By leveraging deep learning techniques, the proposed solution will address current limitations in accuracy and consistency, targeting a 100% visual detection rate for anomalies such as: part […]

Read More
A Scalable Nitrous Oxide (N2O) Emissions Predictor Utilizing Fuzzy Logic and Impulse Response Measurements

Nitrous oxide (N2O) is a powerful greenhouse gas with a global warming potential 298 times that of CO2, significantly contributing to climate change through agricultural activities, particularly the application of fertilizers. Accurate and scalable prediction of N2O emissions is essential for sustainable agriculture and climate mitigation. Current methods, such as LSTM-DLM neural networks proposed in […]

Read More
Understanding Counsellors’ Perspectives on the Mental Health and Wellbeing of Farmers: A Descriptive Qualitative Study

SaskAgMatters (SAM) Mental Health Network Inc. is a support network created to offer a farm-culture friendly mental health support system for Saskatchewan (SK) Agricultural Producers (hereafter collectively called ‘farmers’). SAM offers six free counselling sessions with registered mental health professionals who have backgrounds in agriculture. SAM’s counsellors will be invited to semi-structured interviews focusing on […]

Read More
Airmarket-iART Safety Assessment

Transport Canada is the regulatory agency in Canada that oversees the certification of RPAS (Remotely piloted aircraft systems) and has presented a roadmap for RPAS manufacturers to demonstrate the evidence that a given platform can meet safety objectives. The project will be completed by the iART alliance and aims to develop the technical evidence required […]

Read More
Développement d’une solution d’inspection de quais portuaires par drone sous-marin, photogrammétrie et intelligence artificielle

Le Port de Sept-Îles, premier port minéralier en Amérique du Nord, célèbre 25 ans d’indépendance. Depuis 1999, il mise sur l’innovation, le développement durable et des partenariats solides pour soutenir la communauté. Avec une capacité annuelle de 100 millions de tonnes, il joue un rôle central dans le développement régional et international. Dans ce contexte, […]

Read More