Building for the future, the Department of Electrical Engineering and Computer Science (EECS) in the Lassonde School of Engineering, is creating an AI specialization in its graduate program and will soon offer a professional degree program focused on AI.
Person-following robot to help aging population
Under Professor John Tsotsos (Distinguished Research Professor, Canada Research Chair in Computational Vision and VISTA member), students Raghavender Sahdev and Bao Xin Chen are building visually-guided mobile robotics – specifically, a person-following robot that can absorb visual information, then act on it.
“We built an extensive dataset for person-following robots under challenging situations. We evaluate the proposed system by comparing our tracking approach with existing real-time tracking algorithms,” Sahdev and Chen explain.
As illustrated, this robot uses deep learning in challenging situations, including blockages, changes in target appearance or positioning changes, such as crouching. The video shows how the robot follows the target.
Sahdev and Chen see these robots playing a key role in future eldercare, as they could help seniors by following their charges around from room to room. This would be one of several important functionalities of a companion robot that would enable monitoring for falls and immediate requests for assistance, carrying items, or direct voice communication as one moves about a home or institution.
Person-following robots have many applications, including autonomous grocery carts or personal guides in hospitals or museums. “I believe in working on projects that have applications in the real world,” Sahdev explains.
“I believe in working on projects that have applications in the real world.” ̶ Raghavender Sahdev
Importantly, Sahdev and Chen’s work has received awards at two conferences this year.
Potential for robotics in state-of-the-arts agriculture
Professor Dan Zhang, York Research Chair in Advanced Robotics & Mechatronics and the Kaneff Research Chair in Advanced Robotics & Mechatronics in the Department of Mechanical Engineering, is also interested in robotics to aid society. His main research areas include robotics and mechatronics (technology combining electronics and mechanical engineering), sustainable manufacturing systems, rehabilitation robots and rescue robots.
In 2011, with his Postdoctoral Fellow (PDF), Zhen Gao, he invented a groundhog like rescue robot. More recently, with his PDF Bin Wei, he published Robotics and Mechatronics for Agriculture (CRC Press, Taylor & Francis Group, 2018), which introduces the state-of-the-art technologies in the field of robotics, mechatronics and automation in agriculture.
Other Lassonde researchers are working on on-screen avatars that respond to commands, robocar technology, improved medical diagnostic tools, and visual and motor devices for use by disabled children and adults.
Successful start-up designs custom AI software
Custom software, naturally, plays a key role in AI. Ehsan Fazl Ersi, former PhD student supervised by Tsotsos, has joined with Innovation York to launch a new company, OcularAI, which designs, develops and builds custom AI software. This is Innovation York’s first revenue-generating start-up.
OcularAI’s team of researchers, big data experts and software developers bring a top-notch knowledge base. This enables OcularAI to understand a company’s technical problem, so that highly accurate AI can be created. OcularAI has design capabilities in every aspect of AI, such as computer vision, natural language processing, big data and machine learning.
The services offered by this team also include image/video mining, text mining, social network analysis and scientific discovery.
Since last year, OcularAI has engaged with three Canadian companies.
Thought leadership highlights societal impact of this work
Lassonde researchers also contribute to thought leadership around AI. Professor Marcus Brubaker, VISTA member working in computer vision and machine learning, just wrote an article in Techvibes on the importance of impact in research, specifically in AI and computer sciences.
Brubaker, co-founder of Structura Biotechnology, which applies machine learning techniques to estimating the structure of biomolecules, believes that we’re only scratching the surface of AI’s potential. In the Techvibes article, he speaks about the transformative impact of AI, and urges researchers and practitioners to push ahead on the road to discovery.
“As researchers and practitioners, it’s ultimately up to us to decide how these techniques are applied and to prioritize which applications are most important. We make these decisions and set these priorities with each new project we undertake and, because of the responsibility these judgment calls entail, I believe it’s now more important than ever to question the value and societal impact of the work we do,” Brubaker writes.
“As researchers and practitioners, it’s up to us to decide how these techniques are applied. I believe it’s now more important than ever to question the value and societal impact of the work we do.” ̶ Marcus Brubaker
This is an area where York will shine. Watch this space.
To learn more about York robots, see The York University Magazine article (Fall 2017). For more on Zhang’s work, see his faculty profile. To learn more about his book, visit the publisher’s website. To learn more about Brubaker, see his faculty profile, or read his article in Techvibes, visit the website. To learn more about Structura Biotechnology, visit the website. To learn more about OcularAI, visit the website. For more on the upcoming AI conference, visit the website.
By Megan Mueller, manager, research communications, Office of the Vice-President Research & Innovation, York University, firstname.lastname@example.org