Engineering prof Yimin Yang sees the pursuit of artificial intelligence as a long-distance race that everyone is running.
“Every stakeholder is in the race, they’re just not sure what’s at the finish line,” he said. “Everyone is either interested or panicked about AI technology — or a mix of both.”
Yang lives and works in the middle ground: He worries about the risks of artificial intelligence (AI), especially without more education on how to use it effectively, but he also sees the immense value.
“We also have a duty to teach students to understand these topics. They are the future. If they are aware and taught to harness these technologies, maybe there will be complex problems we can solve in 20 or 30 years.” – Yimin Yang, AI researcher and Western engineer professor
In his research, Yang has seized the technology as a powerful tool for good, using AI to improve everything from house hunting to breast cancer screening.
“My main research focuses on developing new training methods to make AI models more efficient and easier to train, resulting in models with higher cognitive abilities and lower training costs. From an industrial perspective, it’s about helping businesses train more affordable AI models to meet their specific needs.”
He recalled his interest in the technology back in 2006 during his master’s degree in China. Others told him if he focused on neural network research, he’d never be able to get a job.
“Historically, AI has had several winter periods. Seeing ChatGPT become so successful, I don’t think it will enter another downturn. It is very useful; everyone will use it,” Yang said.
“The situation has changed dramatically.”
He’s partnered with Ashirbani Saha, professor and the BRIGHT Run Breast Cancer Learning Health System Chair in the department of oncology at McMaster University. Together, they’re working to improve the reliability and trustworthiness of detecting abnormalities, using AI, in mammograms used for breast cancer screening.
This will support the development of innovative AI-assisted technology for interpreting mammogram results, in close collaboration with a multidisciplinary team. This research is especially important as starting this fall, the Ontario government is connecting more women to breast cancer screening by lowering the eligibility age through the Ontario Breast Screening Program.
Analyzing home photos with AI
Yang also works with a company called Wahi, a Canadian real estate agency, perfecting an AI system to identify the features of a home based solely on its listing photos.
Wahi offers people who are house hunting the real estate equivalent of an online dating app. Swipe on the houses you like, your partner does the same, and voila. It’s billed as the best “house-hunting experience for couples.”
Yang’s work alongside Eman Nejad, the head of data science at Wahi, helps narrow the search even further, using AI to weed out houses without certain characteristics. Search for a finished basement, a kitchen with a good-sized pantry or a renovated primary bathroom.
His team trained the AI model to determine whether a photo depicts a home or room that’s in poor, standard or good condition. They hope to take the work even further in the future, by using listing photos and the information they share to also generate a price prediction for the home.
It makes swiping for houses easier and more pleasing for the users, but the partnership is also a great opportunity for the researchers, Yang said.
“When we have a collaboration with an industry partner, my students and I learn cutting-edge applications and ideas. It’s a bilateral relationship, both parties learn from each other.” – Yimin Yang, AI researcher and Western engineer professor
One PhD candidate and one postdoctoral researcher are also working with Yang, an electrical and computer engineering professor, on the real estate project.
Using AI in an unexpected way, whether it’s medical imaging or looking for a place to live, shows how interdisciplinary partnerships can redefine traditional industries and generate innovative solutions, Yang said.
His goal is always to apply AI to real-life problems. And Yang is proud of turning a complicated, sometimes-feared technology into a helpful tool.
“For me, as a machine learning researcher, I feel the most important and interesting part is that finally, we integrated this AI technology into daily life,” he said. “We are bringing it to the ground, where everyone can use it.”
