Meet the Faculty - Catherine Qi Zhao
Tell us about your journey to the University of Minnesota.
I am currently an associate professor in the Department of Computer Science & Engineering. I first ventured into computer vision and machine learning in grad school at the University of California. At that time, artificial intelligence (AI) was not as capable or well known as it is today. While developing AI, I was intrigued by why machines struggle with tasks that are so trivial for humans. This sparked my curiosity into deeper questions beyond technology and computer science. I became fascinated with the workings of biological systems and the discrepancies between machine and human intelligence. That drove me to study neuroscience at the California Institute of Technology. After my post-doc training, I became an assistant professor at the National University of Singapore where I studied computer attention and broader AI systems. Then in 2016, I joined the University of Minnesota.
I was attracted to Minnesota because of its strong reputation for research in computer science. I also liked the collaboration and inclusive culture, as well the proximity to medical schools and hospitals. I like the vibrant Twins Cities area and the connections with so many local tech and medical companies.
We would love to hear more about your research!
In broad terms, my research focuses on developing and deploying AI systems that assist and augment humans. I firmly believe that AI systems should be geared towards human needs and values. My research seeks to address both the technological hurdles and the alignment of solutions with those principles in mind.
Our technology focuses on providing innovative data and algorithm solutions, and building systems that can see, reason, and learn a variety of real-world tasks and interact with humans. We draw insights from behavioral neuroscience to ensure that our systems are trustworthy and aligned with human behavior and understanding. To that end, we logicalize and improve the decision making process of deep neural networks, leading to the creation of explainable systems. Additionally, we develop new compositional methods and integrate external knowledge just as humans do. We integrate those into models to reduce their dependence on training data and improve their ability to generalize to new situations that the model has not seen before. We work closely with experts to identify critical needs for AI to make a real-world impact.
What do you hope to accomplish with this work? What is the real-world impact for the average person?
I am fascinated by the rapid advancements of AI technology. I think the current AI boom represents a pivotal moment in the creation of machine intelligence that has the potential to impact society more than any other time in human history. I think this is a really exciting time and I am dedicated to contributing to this movement. I want to research or develop systems that have great capabilities, but are also transparent, ethical, and reliable. I want to actively translate those findings into tangible technologies that can benefit other domains like healthcare and manufacturing. The ultimate goal is to develop AI for the betterment of society.
For the past 10 years, I have been collaborating in translational projects to develop AI tools. For example, we developed a novel machine learning model for quantitative, objective, and rapid evaluation of mental health. This work was featured in a Neuron review article and many news outlets for its potential to revolutionize mental health screening. We have also developed deep learning models with neural data to decode human intention. This could potentially benefit millions of people who suffer from movement disabilities.
We also recently partnered with Seagate to pioneer AI-enabled smart manufacturing in response to the national chip shortage in the semiconductor industry. These are a few examples of our efforts to not just develop AI, but to develop AI that will help humans and have a significant impact.
What courses do you teach? What can students expect to get out of your course?
I regularly teach CSCI 5521 - Machine Learning Fundamentals. It is a comprehensive, introductory machine learning course where students learn both the theoretical underpinnings of machine learning and how to practically apply these methods to problems using machine learning and artificial intelligence. This course is mathematically rigorous, so I place a strong emphasis on helping students make connections between theories, visuals, hands-on practices, and applications. In addition to gaining the knowledge set, students leave the course feeling motivated and confident in their ability to use these skills in their future work.
I think teaching is not just about the knowledge set. More importantly, it inspires students to be motivated and confident in their own abilities and take on a growth mindset. They can learn difficult concepts and techniques and apply them. They have the ability to do things themselves and learn and improve.
What do you do outside of the classroom for fun?
I like reading good books. For example, I like “Quest for Consciousness” and “Superintelligence: Paths, Dangers, Strategies”. I find these books very inspiring and thought provoking. They have had a profound impact on my personal and professional growth at different stages.
I also enjoy exploring and experiencing good food, especially from different cultures. It is a fun and delicious way to learn about other cultures and connect with people.
Favorite spot in the Twin Cities?
I really appreciate the lakes in the heart of the city. The lakes and the green space are great for relaxation and exercise. I always feel refreshed after spending time by the lakes. Even driving by them on the way to work is nice.