Visual Computational Sociology: Methods and Challenges
Timnit Gebru (Microsoft Research)
Thursday, 4/12/2018, 10:30 am to 12:00 pm. ARCHIVED EVENT
Location: 1670 Beyster Bldg, 2260 Hayward, N. Campus
In this presentation, Microsoft's Timnit Gebru will describe work intended to increase understanding of a country's demographic makeup, and thereby better target socioeconomic policies.
The US spends more than $1 billion a year gathering census data such as race, gender, education, occupation and unemployment rates. Compared to the traditional method of collecting surveys across many years, which is costly and labor intensive, data-driven, machine learning-driven approaches are cheaper and faster -- with the potential to detect trends in close to real time.
In her work, Gebru and colleagues leverage the ubiquity of Google Street View images of cars to develop a computer vision pipeline to predict income, per capita carbon emission, crime rates and other city attributes from a single source of publicly available visual data.
She will describe this data-mining work as well as other related projects, and discuss how to move the field as a whole toward transparency and accountability.