Doae Kim
Pronouns: She/her
Research Mentor(s): Carol Flannagan
Co-Presenter: Zhang, Ella
Research Mentor School/College/Department: UMTRI / Engineering
Presentation Date: April 20
Presentation Type: Oral5
Session: Session 2 – 11am – 11:50am
Room: Breakout Room 1
Authors:
Presenter: 3
Abstract
In everything from autonomous vehicle testing to improving driver safety, video recordings of drivers provide important data, but to be useable, the data must first be extracted. Computer vision algorithm is a powerful tool that can be built for safety in the market launch and popularization of autonomous vehicles. In an autonomous vehicle, not only drivers but passengers are easy to be vigilant about their surroundings or themselves, so there is a need for solidly constructed technology to develop a safety that can respond to it. My research team works lie in data collection by labeling videos and images of vehicle occupant behavior, as well as objects and actions in the vehicle’s forward view. The data is collected based on three use cases: motion sickness, task engagement, and unusual behavior. Also, there is a potential to introduce an unintended bias in the algorithms that could have negative societal implications. Throughout this research, measuring and reducing this bias such as gender and race bias in artificial algorithms will be a key goal or the big picture of our work. The goal of this study is to develop training datasets for CV algorithms that are reliable and valid enough to make useful and helpful judgments based on people’s behavior.
Engineering, Social Sciences
