What do you do?
I am currently focusing on 3D pose estimation (e.g. animal, human, and object pose) for my Ph.D. dissertation while I am also involved in a couple of vision and language projects such as in food type recognition and communicating public and social media.
Why did you choose this field?
During winter break 2011-2012, I was working on a short-term and intensive project with Professor Christopher Ré (back then at UW-Madison, currently at Stanford) on near-duplicate video detection using both video content as well as the social media signals. I was mentored by Fatemeh Panahi and Ce Zhang and had frequent meeting with Chris and learned a lot of engineering, coding, and technical skills in such a short time. However, it wasn't until later when I worked with Professor Vikas Singh in 2017 that I realized "Computer Vision" is my true calling when I was hacking the Grand Theft Auto V game to control Mike for capturing human pose data while stealing information from the GPU buffer using a Graphic Debugger called RenderDoc. Vikas encouraged me to apply for a CS Ph.D. when I was working as a machine learning intern with him at UW-Madison and apply to computer vision programs. In between starting my CS Ph.D. at BU and my research internship at Vikas' lab I did a short-term Research and Development Engineering role at UC-Berkeley working with Allen Yang and Joseph Menke two experts in computer vision. Their team was very vibrant and I was also called to attend their sister teams' meetings where I was exposed to lots of ongoing and trending research in AR/VR, controlling drones, SLAM, and motion capture. During my R&D Engineer 1 job, I was contributing to test and development of an augmented reality kit, specifically contributing to egocentric hand pose estimation. I was always intrigued by all the things I was learning in CS and even more so intrigued by computer vision, machine learning, deep learning, and natural language processing. I always knew if I have an idea, I can implement it and that made me feel so powerful and in control.
What do you look at and think, "I wish younger me would have known this was possible"?
When I was a CS master's student at UW-Madison, I always wanted to become the President of Women in ACM chapter. However, there were always way more qualified women that I even voted for them becoming a president. Later when I joined BU, I realized we don't even have a official ACM-W chapter. I definitely felt it was missing and always remembered how the Women in ACM chapter helped me decide to start a master's in CS in 2014. It also made me become friends with many women in our department whom I wouldn't know about otherwise. It took me two years of not taking any action to one day telling my adviser, Professor Margrit Betke, that I want to start this club if she think it's ok to do so as a Ph.D. student given all other workloads I have. She was very supportive of my idea and connected me with Ziba Cranmer, director of Spark! at BU who later jointly helped me co-found the BU ACM-W in November 2019 as the President. I never thought I would found a club and never would have known how much work it is but I am truly very proud of myself and would tell younger me, believe in your dreams and your dreams coming true one day or another. Maybe not exactly in the same form you want them to be. What I love about ACM-W is this bond we have right now between undergrad, master's, and Ph.D. female students. Something that I personally felt missing during the first two years of my Ph.D.
Why do you love working in STEM?
Working in STEM, given my love of computer science, mathematics, reasoning, and logic, I never felt I am working. For the most part, I was always enjoying what I was doing. I always felt I am working on a puzzle except maybe I didn't have the exact big picture or sometimes I only had a hallucination of it. Still when I was putting these puzzle pieces by solving all these interconnected smaller problems towards a goal, I was getting a lot of satisfaction every time I got to see a local map of the big picture. Recently, I told my mom that I was awake at 3am working on some code and she told me I have no idea how you could keep awake otherwise if you didn't love CS. Sometimes, I don't even understand the hours go by. I think everyone should do what they truly are passionate about. From personal experience I can tell that it's not an easy task to figure what you are truly passionate about from the get go and it takes a lot of rounds sometimes to find your global maximum. I have been personally stuck in local maximums which at the time I felt I love, and then realized about a new research field and then moved to a better global maximum but currently I feel I am in a global maximum and have found my calling, computer vision research. Recently, I've been hooked up to this song "Gasoline" by Cash Cash which sums up my true feelings for computer vision. "There's no end to you and me. It's like putting out, fire with gasoline". Like I am the fire and computer vision is my gasoline. I get energized when I work on a new problem. Maybe you can look at yourself and see how you feel. Do you feel tired after an hour of working or do you feel empowered after an hour? That could help you decide if you are following your passion or not.
Best advice for next generation?
I am really glad that in the USA there is a lot of initiative for middle school and high school students including girls to follow engineering and science. I think this should happen worldwide and it potentially could start earlier. If we can learn math or science in elementary school, I bet we can learn computer science at the same time too, maybe using some platform like Lego robots or MIT scratch.
Inspo quote / fun fact / role model
"Find what you love and let it kill you" ~Charles Bukowski