CSpotlight: Journey of a computer vision engineer

Recent M.S. graduate Pallavi Mitra discovered her passion for autonomous systems as an undergraduate in India. She was heavily inspired to continue her studies at the University of Minnesota thanks to the research being conducted by Computer Science & Engineering (CS&E) professors Junaed Sattar and Hyun Soo Park in the field of computer vision. When the pandemic hit during her last semester, she was able to remain motivated, finish her thesis, and land a job as a computer vision engineer.

Tell us more about your research thesis. What inspired your interest in monocular depth estimation?

I became fascinated by the way autonomous systems work as an undergraduate student. After coming to the United States to begin my graduate studies, I slowly started learning about the technology behind autonomous robots. Through my coursework, I started to take an interest in the working of self-driving vehicles, underwater autonomous systems, and other space constricted autonomous systems. Monocular depth estimation is immensely useful for autonomous systems where space constraint is a crucial factor. I was inspired by the working of such systems and wanted to contribute to the technology behind it. I found it extremely interesting and challenging to deal with this problem.

How did you become a part of the Interactive Robotics and Vision (IRV) Lab? What was the highlight of your time in the lab?

I joined the IRV Lab in spring 2019 for an independent study coursework. Later, I chose to pursue a Plan A (the thesis-based program) master's degree with focus on monocular depth estimation, with Professor Sattar as my advisor. The research experience in the lab enhanced my thirst for research and helped shape me into an individual capable of making an impact in the computer vision research community. Needless to say, my confidence and desire to apply my technical knowledge to solve real world problems increased significantly during my stay in the lab. I am grateful to Professor Sattar for providing me with this opportunity.

What has been the biggest challenge you’ve faced because of the pandemic?

Adapting to “work-from-home” was quite challenging for me, especially during the beginning of the pandemic. As a social person, the reduced human interaction and communication, coupled with the Minnesota winter, affected both my mental and physical health. Also, being a part of the queer community, socialization has been even more difficult.

Tell us about your work as a Computer Vision Engineer at Sentera. What kind of projects are you working on? 

Sentera works in the domain of precision agriculture, building drones and producing insights on the health of crops to provide recommendations for farmers. I joined the company in September 2020 and have been privileged to work on some exciting projects on weed detection and data quality control. Working remotely has not been challenging thanks to my amazing co-workers, who made the experience very interactive and seamless.

How did your experience in the M.S. program prepare you for your career?

Pursuing my computer science M.S. at the University of Minnesota has made me very confident in my technical knowledge and exposed me to related fields like virtual reality and computer vision. Having experience as both a teaching assistant and a researcher also played a significant role in my journey to become a computer vision engineer.

What advice do you have for others considering the U’s computer science master’s program?

Prospective students should communicate with professors and investigate the wide variety of course and research options available at the University. Many faculty members, like Professor Sattar, are extremely supportive of student researchers in his lab. I would also encourage students to be proactive and actively engage in the leadership and development opportunities that are available through the CS&E department, including the Grace Hopper Conference, which I was fortunate enough to attend in 2019.