Unsupervised Anomaly detection using Deep Learning

This project aims at evaluating whether recent results in deep learning models, trained to exploit weak labels can serve to extract meaningful lesion localizations from image-level labels, either from individual scans or given a (longitudinal) sequence thereof. To this end, we will scale up existing models that have been shown to work on 2D images […]

Read More
Improving Efficiency and Robustness of Model-based Reinforcement Learning

Model-based reinforcement learning allows AI systems to learn and use predictive models of their environments to plan ahead, achieving tasks more efficiently. The proposed project aims to (i) develop methods for identifying when an uncertain and/or flawed model can be relied on to make plans, and when it cannot, and (ii) implement a method which […]

Read More
Integrating Inclusive Design into Project Workflows

The World Wide Web Consortium (W3C) has published a set of guidelines to assist designers in the task of designing digital content which will be accessible to users with a wide range of abilities. These guidelines are known as the Web Content Accessibility Guidelines (WCAG), version 2.1. Students in the Inclusive Design program learn to […]

Read More
Wide-baseline Novel Scene Synthesis from a Single Image

Novel view synthesis is the process of generating new images from an unseen perspective, given at least one image of a scene. There may be more than one probable novel view associated with each unseen perspective, an assumption made by existing methods. This simplifying assumption prevents these methods from being applied to more difficult novel […]

Read More
Automatisation de la classification par magnétométrie quantique de failles en présence d’erreurs de mesure

De nombreuses structures sont enfouies sous terre dont les conduites de transport de liquide dangereux tel que le pétrole. Lorsqu’une alerte est donnée pour indiquer qu’il y a possiblement une faille dans une zone de la conduite, identifier l’emplacement de la faille nécessite soit de déterrer la conduite, soit l’utilisation de magnétomètre (détecteur de champ […]

Read More
Investigating multi-task learning in semantic parsing

Current research in semantic parsing suffers from lack of annotated data, which is hard to acquire. In this project, we aim at tackling the problem of converting natural language utterances to SQL language (Text-to-SQL) on complex databases in a low-resource environment. More specifically, we focus on the research of how multi-task learning (MTL) can help […]

Read More
POLITICAL ADVOCACY BY FOR-PROFIT ENTERPRISES: ACTIONS AND GOVERNANCE IMPLICATIONS

For profit companies can exercise influence over public politics: Advertisers withdraw funds from media programs and media outlets, companies move service centres and offices, companies stop providing services to certain clients. This research will perform a series of comparative case studies, using publicly available commercial communication and interviews with industry professionals to understand how for-profit […]

Read More
Measuring entanglement in quantum magnetic systems with strong long-range correlations

D-Wave systems purports to have designed a quantum processor based on scalable architecture that physically implements quantum annealing, an algorithm that can be used to solve a wide variety of optimization problems. In order for D-Wave devices to exhibit a performance advantage over classical processors, it is necessary that the devices utilize a resource that […]

Read More
Formation of Supermassive Black Holes via merger of Protostars

The origin of Supermassive Black Holes that are observed to exist in the early Universe is one of the key questions in astronomy. There are different models that try to explain how these black holes formed and grew in a short period of time. One such model is the collision of protostars resulting in massive […]

Read More
Health Records over Blockchains

It is hard to understate the critical role access to a patient’s comprehensive medical history plays in diagnosing (and treating) patient illness. For a doctor, knowing the latest prescribed drugs of his patient might, for example, point directly to the cause of an illness — which might be just a drug side effect. Yet patients […]

Read More
Health Records over Blockchains – Year Two

It is hard to understate the critical role access to a patient’s comprehensive medical history plays in diagnosing (and treating) patient illness. For a doctor, knowing the latest prescribed drugs of his patient might, for example, point directly to the cause of an illness — which might be just a drug side effect. Yet patients […]

Read More