Indirect Domain Shift for Single Image Dehazing

Deep convolutional neural networks (CNNs) have been tremendously successful in many high-level computer vision tasks, e.g., image recognition and object detection. Although recent works have shown that it is also possible to learn an end-to-end CNN model for low-level vision tasks, e.g., image dehazing, the resulting performance is still not completely satisfactory. For high-level vision […]

Read More
Quantitative biomarkers from Magnetic Resonance Imaging

Magnetic Resonance Imaging (MRI) is the non-invasive method of choice for diagnosing and studying neurodegenerative diseases such as multiple sclerosis (MS). However, conventional MRI that is currently used in clinics cannot provide a reliable measure of neuronal health. More specifically, it cannot distinguish healthy from diseased myelin, a key component of neuronal tissue that is […]

Read More
Development of Efficient Methods Preprocessing Large Lidar Data Sets for Application to Road Design and Optimization

Technological improvements, competition in the survey services industry and the increased use of UAV’s (drone) has driven down the cost of LiDAR acquisition. As a result, LiDAR is rapidly gaining popularity in application in road planning and design. LiDAR data sets typically contain tens of millions of points. Efficiently processing this data efficiently presents challenges […]

Read More
Speech Localization and Recognition for Humanoid Robotics

Robots and other autonomous machines currently have limited sound awareness and speech recognition capabilities, which limits their ability to interact with humans. Applying insights from cognitive neuroscience provides novel ways to improve these interactions. The primary objective of this proposal is to implement an auditory AI for robotics that finds human talkers in the environment […]

Read More
Decoupling of the structural connectome in the elderly

In 2010, approximately 35.6 million people suffered from dementia around the world and the number is expected to double every 20 years. These alarming predictions call for a better understanding of the mechanisms underpinning cognitive decline in aging. One of the key cardiovascular risk factors for cognitive decline is arterial stiffening. My project involves the […]

Read More
Prédiction de maladies génétiques à partir des forêts aléatoires et régressions logistiques

Le projet consistera a développer des modèles prédictifs utilisant les données de tests médicaux de patients. Quatre algorithmes en utilisant les méthodes de régression logistique et de forêt aléatoire seront utilisés prédire le diabète, l’hémoglobinopathies, le beta thalassémie, le niveau élevé de LDL-C. Les algorithmes permettront de classer divers résultats et d’en déduire un recommandation […]

Read More
Low Data Drug Discovery

The project aims to facilitate the research and development of new drugs by exploring Machine Learning methodology useful for both the generation of new molecules and the prediction of molecule properties. Doing so will involve training deep learning models on a large number of small, heterogeneous datasets, with the objective of transferring learned representations quickly […]

Read More
Improving Receipt Classification Through Text Processing

10sheet is a third party software company aimed at providing easy bookkeeping to client companies. One of the steps in bookkeeping is classifying receipts based on their content. Without relying on a human bookkeeper, 10sheet uses a classification algorithm to build an automatic classifier for this purpose. However, such a classifier needs to have a […]

Read More
Intelligent systems for social community identification and applications

Social Listening is the act of accessing the many available social networks that exist throughout the world wide web. The partner organization of this research has identified algorithms using existing web technology to evaluate and categorize an interpretive community through the connections of that community?s participants and the context of those connections, including both verbal […]

Read More
Fostering Learning and Professional Development for Users of Feature-Rich Design Software

Feature-rich design software is of critical importance to many industries, including manufacturing, architecture, and construction. A common feature of this software is that, to be used effectively, the user must possess technical competency and an understanding of the specific workflows and practices for using the software within their organizations. In the past, users could get […]

Read More