Platform Prototyping for Haptic Steering Research

In “shared control” of a machine, a user works cooperatively with an intelligent system to continuously direct or steer the machine’s operation. In this project with Spark Robotics Technology Inc., a provider of products and services for the defense and security industry, the research team will apply this concept to the problem of navigating or searching information streams.

Machine Translation for Social Gaming

This project with Koolhaus Games, an electronic game developer, aims to use state-of-the-art natural language processing technology as part of a novel and innovative social networking and social gaming engine. Social networking games are popular, particularly among children, and the team hopes to use natural language processing technology to provide a transparent, multi-lingual interface for communication between children from various countries who speak different languages.

Semi-Automatic Video Motion Capture: Lifting 2D Estimates to 3D

Motion Capture (mocap) is a very powerful technology that allows one to digitize natural motion and use it in many areas. Currently, mocap is expensive and intrusive. The core of this project is to develop a pipeline for semi-automatic extraction of 3D motion data from ordinary video sequences. Besides being able to avoid many of the expensive and time consuming aspects of traditional motion capture, the intern will be able to extract motion data in a non-intrusive fashion from readily obtainable video data.

Mapping the Human Kineome and Phosphoproteome

This internship is the first in a series of four which Kinexus is undertaking with ACCELERATE BC which will support its overall ambitious goal of creating high resolution maps of internal cell communication networks that are gender, age, tissue and disease-specific. These maps will be needed for identifying disease diagnostic biomarkers that are phospho-sites and drug targets that are kinases. The dividends of these efforts will be improved diagnostics and therapeutics for personalized medicine delivery.

Analysis and Quantification of the Textural Information in Pigmented Skin Lesions

There are certain features of skin lesions or moles that are indicative of cancer. Texture (spatial variation of colour/intensity) or patterns is known to be one of these features. This research seeks to quantify this textural information in order to improve the performance of methods to automatically diagnose melanoma. Currently, dermatologists are very proficient at interpreting this textural information, and easily outperform automated techniques.

The Gem Cutter Educational Tool

This project will explore the pedagogical usefulness of a visual programming tool developed by Business Objects, called Gem Cutter. It will be used to illustrate and teach computer programming concepts to new students. This is different from the traditional text-based model in which students are required to engage in text-based interactions with a computer as they develop programs. The text-based model tends to be difficult for new students to grasp and often discourages them in early stages of learning.

Roof Plane Generation Algorithm for Slope Roofs

Automated Systems Research (ASR) has been supplying software to the roofing estimation market for close to 20 years. While the lack of competition in this niche market has provided for a steady ongoing business opportunity, it has created a vacuum of research in this area. One of the recurring research problems at ASR is that of generating the entire surface of a sloped roof given the building’s polygonal outline and also the roof slope angle at each outline edge.

Algorithms for Nonlinear Inversion in Medical Imaging

Non-invasive imaging techniques using near-infrared (NIR) light yield extensive information on cells and tissues and can reconstruct images of the deep interior of biological tissues. The place occupied by NIR imaging within the multidisciplinary field of molecular imaging has been growing in recent years. This is, in pat, because of its intrinsic advantages (such as its low cost and non-invasiveness) when compared to other imaging modalities such as positron emission tomography (PET), magnetic resonance imaging (MRI) and computed tomography (CT).

Algorithm for Robust Spam Blog Detection

In the world of today’s internet, ‘web notifications’ are becoming increasingly common. As opposed to a traditional search which is executed once against entered keywords, a web notification search is performed on a regular or continual basis, keeping track of available information changes in real time. The user is then notified once an item of interest appears on-line. Notification services receive continuous pings from blogs which inform of updates.

Features-driven Multi-modality Registration and Fusion of Functional and Structural Medical Images

The internship project with Pacific Nuclear Medicine will develop a software tool that brings two images obtained from different imaging modalities into spatial alignment (aka multi-modal registration). This is a necessary task for the creation of correct image fusions, which are essential for diagnosis and correction of nuclear medicine images. Additionally, the project will develop a feature-registration framework in which pattern-recognition techniques and feature-extraction methods to increase the accuracy and robustness of the registration algorithm will be employed.