Rich Recommendations

Item-to-item and user-to-item recommendations are prevalent on most ecommerce websites and digital content related mobile applications. At Kobo, we strive to constantly improve our recommendation system, which is based on co-purchase patterns on Kobo’s website or through Kobo eReaders and mobile apps. This internship is to explore improving the system along several dimensions: incorporating additional […]

Read More
Business Intelligence and Recommendation for Samsung Knox Community Forum

Companies usually use private forums to share sensitive information and discuss strategies. Valuable business knowledge is hidden inside the user generated content in these forums and waiting to be extracted and reused. Knox Community Forum, a thriving private forum inside Samsung where employers from all around the world post content like win notices of deals, […]

Read More
Automatic Spam Alert System for Online Social Media

Online discussion forums are examples of social media that are becoming popular. The goal is to develop efficient and scalable techniques that work on large amounts of conversational text to separate the posts that truly matter from irrelevant ones, as well as techniques that are easily adaptable to new domains and forums. VerticalScope, Inc. is […]

Read More
Human Activity Analysis in Sports Videos

Automated human body pose estimation and activity recognition in videos is still one of the challenging problems in computer vision. Generally, it is becomes a significantly difficult task in real world applications due to camera motion, cluttered background, occlusion, and scale/viewpoint/perspective variations. Moreover, the same action performed by two persons can appear to be very […]

Read More
Dynamic Light Redistribution Algorithm Advancement for High-Dynamic Range Digital Projection

This project will investigate and develop advanced algorithms for high dynamic range projection using caustic image formation. These algorithms dynamically redirect light from dark image regions to bright regions, which addresses the problems of low peak luminance and poor black levels in current projectors. The algorithms will improve upon existing lensing methods by increasing computational […]

Read More
Analyse comparative d’une formation en présentiel et d’une formation basée sur un jeu sérieux dans le cadre d’Hydro-Québec

Depuis quinze ans, Hydro-Québec organise des formations sur l’Initiation au code de sécurité des travaux Postes et centrales et Initiation aux risques d’induction, mise à la terre et foudre. Ces formations de trois jours s’adressent à toute personne appelée à intervenir chez Hydro-Québec. Visant un meilleur coût d’opportunité, Hydro-Québec a choisi d’intégrer une nouvelle modalité […]

Read More
Can Learning Catalytics Replace Clickers?

Classroom response tools such as the ‘clicker’ are devices that students use in class to answer multiple-choice questions. Many studies have shown that clickers can lead to increased learning when properly used. Learning Catalytics is a new web-based tool marketed by Pearson that is meant to replace clickers. While Learning Catalytics offers more features than […]

Read More
Examining the construction of masculinity in K-Dramas within a global context

This research project analyzes how South Korean TV dramas, particularly King the Land (2023), portray masculinity and why these representations appeal to Western audiences like those in Canada and Brazil. By studying cinematic techniques and cultural themes, the project aims to understand how these dramas shape global perceptions of South Korean men and culture. The […]

Read More
Estimation de la pose de la caméra physique en production virtuelle

Cette recherche explore l’intégration des espaces physique et virtuel en production cinématographique. L’espace physique regroupe acteurs, décors, éclairages et caméras, tandis que l’espace virtuel crée des environnements numériques immersifs. Pour un rendu visuel réaliste, les mouvements de la caméra physique doivent être précisément suivis et reproduits dans l’espace virtuel. Actuellement, ce suivi repose sur des […]

Read More
HPN Hyperautomation Platform

HPN is building a hyperautomation platform using data science, artificial intelligence (AI), and machine learning (ML) techniques to hyperscale its unique business model. The expected outcomes from this platform will be scale up of revenue internationally along with improved operating margins; enhanced member and customer experience through self-service and AI concierge services; and innovation in […]

Read More