Dr. Lynne L. Grewe phone: 510-885-4167, Room SF 551 Dr. Grewe has her PhD from the Electrical and Computer Engineering Department of Purdue University. She has a number of years industrial experience and academic experience. Her research interests include Computer Vision, Media Processing and AI. She has recently served while on sabbatical at Google as a Faculty in Residence.
|
Tell about your foray into tech world?
- I began my career as an Undergraduate student as an intern at Intel Corporation. It was exciting to come to Silicon Valley, the heart of much tech invention and becoming part of a major leading tech company. I learned a lot including what I didnt' like to do. I strongly encourage students to look for any opportunity to work/participate in the tech industry before they graduate.
- I subsequently worked at HP and Hughes Research Labs. I learned working at a research lab how much I liked the independence and autonomy of doing research.
- I worked full time at IBM corporation who also sponsored me for my PhD. I was a Vision specialist and worked with research and an International team on unique vision based systems. I enjoyed working on an international team but, working across time zones is a challenge.
How do you use Web Systems in your work?
- Besides teaching, I do research in Computer Vision and Machine Learning. As part of this I sometimes create user interfaces, back end systems and use various web services for both development and production.
- Today, most computer vision and ML system s require the use of ML frameworks and involve relatively high computationally expensive training. For this I use a range of Cloud (web) services.
- As part of that work I create prototype systems. These systems may have web or mobile interfaces.
- Often I have need to integrate my prototypes with web services. Consider the "Blind Bike" system that helps low vision people with the task of biking. Part of this is navigation and the use of maps as well as turn-by-turn navigation.
- Sometimes, my systems need to store data in the cloud. Consider the "Covid-ID" system which allows crowd-sourcing of images that can be processed to detect safe mask wearing and crowd distancing, etc. These resulting reports need to be stored in the cloud so that users of the "Covi-ID" app can be aware of the situations in difference locations. This means that not only do I need to recieve this data with backend programs, I need to store the data in the cloud (web).
=