- Tuesday, August 4, 2020 - 1:30pm
- See email announcement for Zoom location
Committee:Dr. Mayank Goel (HCII), ChairDr. Jodi Forlizzi (HCII)Dr. Scott Hudson (HCII)Dr. Thomas Ploetz (Georgia Institute of Technology)Abstract:We have entered a new age of computing where the computer is tied not only to a person's body but may also be present in their environment. The ambient presence of sensors enables unprecedented opportunities to build smart environments that adapt to user needs, tracks activities, enables interactions and assists the user in their daily tasks. Even though ambient sensing exists, its scale until recently was limited by the available hardware and computing power. Lastly, ambient sensing techniques tend to be privacy intrusive.
In this dissertation, I identify key challenges of sensing at scale- the ability to track multiple users and activities in an environment at the same time. The need to build reliable novel approaches that detect multiple activities from the same sensor stream and identify the user performing those activities presents a unique technical challenge. Whereas managing the privacy expectations of all users in a shared environment is a socio-technical challenge that influences the design of those approaches. I built two systems that tackle these issues. They demonstrate (1) the ability to sense multiple activities (e.g., different exercise types) for 100s of users at the same time, and (2) the ability to identify individual users in a group. I have also conducted evaluations in unconstrained environments to underscore the practicality of these approaches. Finally, I also outline how both systems take a privacy-by-design approach. I further my work in practical privacy preserving sensing at scale and propose future research directions to better understand privacy expectations of users, and aid developers by improving methods to collect and label data for machine learning algorithms.Document Link: http://rushil.fyi/PhD_Proposal.pdf
- Queenie Kravitz