Promise Into Practice: Application of Computer Vision in Empirical Research on Social Distancing

Social scientists increasingly use video data, but large-scale analysis of its content is often constrained by scarce manual coding resources. Upscaling may be possible with the application of automated coding procedures, which are being developed in the field of computer vision. Here, we introduce computer vision to social scientists, review the state-of-the-art in relevant subfields, and provide a working example of how computer vision can be applied in empirical sociological work. Our application involves defining a ground truth by human coders, developing an algorithm for automated coding, testing the performance of the algorithm against the ground truth, and running the algorithm on a large-scale dataset of CCTV images. The working example concerns monitoring social distancing behavior in public space over more than a year of the COVID-19 pandemic. Finally, we discuss prospects for the use of computer vision in empirical social science research and address technical and ethical challenges. [ FROM AUTHOR] Copyright of Sociological Methods & Research is the property of Sage Publications Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full . (Copyright applies to all s.)

How common are high-risk coronavirus contacts? A video-observational analysis of outdoor public place behavior during the COVID-19 pandemic.

Epidemiological evidence and recommendations from the World Health Organization suggest that close face-to-face interactions pose a particular coronavirus transmission risk. The real-life prevalence and nature of such high-risk contacts are understudied, however. Here, we video-observed high-risk contacts in outdoor public places in Amsterdam, the Netherlands, during the COVID-19 pandemic. We found that high-risk contacts were relatively uncommon: Of the 7,813 individuals observed, only 20 (0.26%) displayed high-risk contacts. Further, we qualitatively examined the 20 high-risk contacts identified and found that they occurred disproportionally between affiliated persons engaged in affiliative behaviors. We discuss the potential public health implications of the relatively low incident rate of high-risk contacts.