The study sampled 98,000 street and walking-route locations in 98 cities and obtained 9,997 Google Street View images from 82 cities. Among sightings the model could classify, women accounted for 28.8% of pedestrians and 5.4% of two-wheeler riders. The model could not classify 66% of all sightings.
Cite: Gaurav Sood. "Measuring Women’s and Men’s Presence on Urban Streets: Evidence from Google Street View in 98 Cities." https://gsood.com/research/google-street-view-98-cities/
BibTeX:
@misc{soodgoogle,
author = {Gaurav Sood},
title = {{Measuring Women’s and Men’s Presence on Urban Streets: Evidence from Google Street View in 98 Cities}},
note = {Working paper},
url = {https://gsood.com/research/google-street-view-98-cities/}
}