Áù¾ÅÉ«ÌÃ

Our research group is currently engaged in several projects applying machine learning with industrial applications (Baker Hughes, ROSEN group, City of Calgary). We are developing deep learning-based techniques in anomaly detection in video, remote sensing imagery, magnetic flux, and acoustic sensors. Hybrid data-driven modeling approaches and integration of physics, cyber-physical data are used to understand the influencing factors of a system and predict behaviours given sensed data in applications of fluid flows in pipelines, stock markets and power systems.

We have developed and tested a persistent acoustic monitoring system and 3D people-counting system with edge analytics for City. We are developing novel approaches for robotics and 3D sensing for automation in outdoor and indoor environments. In machine learning, we are pursuing the use of spiking neural networks (SNN) and implementations on FPGA/embedded systems for low-power edge computing applications above. We are also investigating large-language models and the capability of semantic understanding. We are developing an AI framework for trustworthy situational awareness (Department of National Defence IDEaS project). We are currently developing techniques for AI processing aimed at IoT and UAV applications.Ìý

We worked on distributed sensor fusion development. Our developed multi-sensor registration and fusion algorithms based on expectation maximization have been successful in simultaneous registration and fusion of dissimilar sensors. We also work on the higher level information processing for data fusion where Bayesian networks are utilized for situation assessment. Another challenge in sensor network is the limited bandwidth available for communication. Our analog wideband communication schemes aim to achieve high data rate at low transmission power levels.

One of the achievements of our team in this area is the application of nonlinear systems in analog wideband communication. The introduction of the ergodic chaotic parameter modulation scheme had been a significant contribution, towards this end. We have a special interest in target tracking using the radar. Our team has developed advanced signal processing techniques combined with the power of nonlinear dynamics to perform through-the-wall imaging. Chaos signals have been successfully used in spread spectrum as well as time modulated ultra wide band radars. We developed a chaos based receiver to perform blind equalization for room reverberations. When intelligent robots are applied for the surveillance, navigation control becomes an important problem. Our research team also worked on the cooperation of multiple robots to achieve faster and efficient coverage using chaos synchronization and genetic algorithms.

Research Topics

1. Anomaly Detection and Industrial Applications

2. 3D sensing and Robotics

3. Machine Learning and Edge Analytic

4. Internet of Things (IoT) and Signal Processing
Ìý

IDEaS Micronet

A collaborative project involving innovation from 5 Universities across Canada and 10+ Industry partners, combining 5G, Artificial intelligence, Edge analytics, UAV robots for enhanced situation awareness. The project title is: 5G-Enabled Trustworthy Common Operational Picture with Edge Server Data Engine (5G-TCOP). Click here for more information.

Opportunities

Information for Prospective Students

We are currently looking for outstanding MSc and Ph.D candidates. Students particularly interested in the areas of Communication Theory, Statistical Signal Processing, Information Fusion, Chaos & Nonlinear Dynamics, Intelligent Systems, Machine Learning and Robotics but whose research interest does not fall under any one of the ongoing projects may submit a brief research proposal along with their application for review. The "Autonomous Systems and Intelligent Sensing Laboratory" offers a broad range of lab facilities including robot development platforms, communication FPGA kits and computing facilities, and access to computing clusters with the University. Interest candidates may send an email clearly stating your research interests, prior education & training relevant to the research and list of any publications related to the proposed research topic. Accepted students are guaranteed financial support for a period of 36-48 months for Ph.D and 24-28 months for MSc programs. The candidates are expected to be proficient in one or more programming languages and have strong mathematical and theoretical background. Accepted students need to attend weekly group meetings to discuss their research progress.

Information for Post-Doctoral Fellows & Visiting Scholars

We are currently looking for outstanding candidates for the position of postdoctoral fellows. Applicants who have conducted research in the areas of Communication Theory, Statistical Signal Processing, Information Fusion, Chaos & Nonlinear Dynamics, Intelligent Systems, Machine Learning and Robotics and have a strong publication record are encouraged to apply for the position. Interest candidates may send an email clearly stating your research interests, curriculum vitae and a synopsis of their PhD thesis.

Ìý

News

Oct 2024:

May 2024: A snowy start to the new spring semester! The website has been updated.

Jan 2023: Happy New Year! A warm welcome to the new students to our lab, and a congratulations to those who completed their graduation requirements in 2022. The current team and website has been updated.

2022:

2022: The website has been updated and redesigned. Pages are in the progress of being updated.

2018:

Archived News

The archived website and news is available:

Ìý