Jonathan Hou


Pleora Technologies

Solving the Deployment Challenge – Moving AI from Research to the Factory Floor

Bio:Jonathan Hou is President of Pleora Technologies, a leading supplier of AI and sensor networking solutions for the industrial automation, security & defense, and medical imaging markets. In this role, Jonathan oversees Pleora’s research & development efforts and leads the company’s long-term technology vision. Jonathan joined Pleora as Chief Technology Officer in 2018. Previously he was Director of Technology with GlobalVision, where he helped develop new automated quality inspection solutions for print inspection applications. He has held positions in software & engineering management, applications engineering, and software development in the machine vision, video, graphics and networking industries. Jonathan has a Bachelor of Applied Sciences – Computer Engineering from the University of Waterloo in Waterloo, Canada, and a Master of Engineering from McGill University in Montreal, Canada.

Abstract: This presentation will separate hype versus reality when it comes to real-world AI deployments. For many manufacturers investigating AI there is still a great deal of confusion around the capabilities of the technology based on “best case” scenarios. This often leads to cost concerns, delayed deployments, and disappointing results. Instead, a more scaled approach that starts with digitization and evolves towards wider AI-based automation often delivers better outcomes for manufacturers. As part of this discussion, we will look at how an electronics manufacturer and distillery are using machine vision and AI-based tools to first digitize error-prone human processes and gather data to help guide and simplify an AI automation strategy.

Rikke Gade

Associate Professor

Aalborg University, Denmark

Thermal imaging for privacy preserving surveillance applications

Bio:Rikke Gade is currently employed as Associate Professor at Aalborg University, Denmark, in the Visual Analysis of People Lab. She received her M.Sc. and PhD degrees from Aalborg University in 2011 and 2015, respectively. During her studies she has also visited University of Auckland, New Zealand and University of Adelaide, Australia. The PhD thesis in Computer Vision focused on analysis of activities in sports arenas; mainly occupancy analysis, activity recognition, and tracking of players. Most of her work revolves around the use of thermal video, to preserve privacy in public sports facilities. This has led to publications in top journals and international conferences within computer vision, and she has been co-organizing workshops at CVPR, ICCV, and ACCV. Her research interests include computer vision analysis of human activities, applied both for analysis of human behaviour at public spaces as well as for analysing sports activities. She also works with robot vision.

Abstract:In this talk I will dive into our story of using thermal cameras for privacy preserving computer vision algorithms at Visual Analysis and Perception lab at Aalborg University. My first encounter with thermal imaging was back in 2011 when thermal cameras were rarely seen in public computer vision research. As part of my PhD project I captured and analyzed long-term thermal datasets of a variety of human activities in both indoor and outdoor environments. The majority of our work revolves around sports applications such as occupancy analysis of sports arenas (indoor and outdoor) and analysis of sports activities where thermal cameras are used instead of regular RGB cameras to preserve privacy on public facilities. The computer vision methods applied ranges from low-level image processing and machine learning to deep learning on our more recent work. Other applications I will cover in this talk include clothing level estimation in office environments and detection of accidents in open harbour areas. During my talk I will, among other things, share insights on what to pay special attention to when using thermal cameras and demonstrate how methods designed for RGB images can be adapted to the thermal domain.

Robert Laganiere


University of Ottawa, Canada

Perception from Radar sensors: principles and challenges

Bio:Robert is a professor at the School of Electrical Engineering and Computer Science of the University of Ottawa and the CEO of Sensor Cortek, a startup company developing AI solutions for perception systems. Robert is the co-author of several scientific publications and patents in content-based video analysis, visual surveillance, embedded vision, driver-assistance and autonomous driving applications. Robert authored the OpenCV2 Computer Vision Application Programming Cookbook (2011) and co-authored Object Oriented Software Development (2001). He co-founded Visual Cortek in 2006, an Ottawa-based video analytics startup that was later acquired by iWatchLife in 2009. He also co-founded Tempo Analytics in 2016 a company proposing Retail Analytics solutions and founded Sensor Cortek inc in 2018.

Abstract: Radar is one of the essential technologies that enables machines to perceive and interact with the world around them. By sensing the environment using radio waves, radar provides information about a scene and its objects that can be used in a wide range of applications, from autonomous driving to surveillance and security. Radar is an old technology that continues to evolve and that is expected to play an important role in the perception systems of the future. In this presentation, a survey of the radar technology and its benefit will be provided. We will explain how radar can extract range, azimuth and velocity information to detect objects and their speed. We will also explore some of the recent development that uses AI to improve radar perception and discuss the particular challenges that pose radar data in the context of deep learning.

Zheng Liu


The University of British Columbia, Canada

Towards a Comprehensive Perception: Methodologies for Thermal Imaging Data Analysis

Bio: Zheng Liu is a professor at the Faculty of Applied Science of the University of British Columbia (UBC Okanagan). Before joining UBC, he worked for the National Research Council of Canada as a research officer and for the Toyota Technological Institute (Nagoya) as a professor. His research interests include digital twin, data/information fusion, computer/machine vision, machine learning, smart sensor and industrial IoT, and non-destructive inspection and evaluation. Dr. Liu is a fellow of SPIE and a senior member of IEEE and holds Professional Engineer licenses in both British Columbia and Ontario. In addition, Dr. Liu serves on the editorial boards for journals including Information Fusion (Elsevier), Machine Vision and Applications (Springer), IEEE Transactions on Instrumentation and Measurement, IEEE Transactions on AgriFood Electronics, and IEEE Journal of RFID, and CAAI Transactions on intelligence technology.

Abstract: Many industrial sectors and applications have benefited from thermal imaging technology for its perception capability enabled by its spectrum and advances in computational methodologies in varied situations. A thermal imaging system can be configured or operated in different modes, e.g., unimodal and multimodal, with or without explicit fusion operations. However, what thermal imaging contributes to human perception depends on the forms of the derived information, i.e., how the thermal imaging data are processed with the auxiliary data and information. This talk will overview the methodologies for processing and analyzing thermal imaging data in the context of application needs. The research opportunities and challenges will be highlighted in the presentation.