RT Article
T1 Explaining Crime Diversity with Google Street View
JF Journal of quantitative criminology
VO 37
IS 2
SP 361
OP 391
A1 Khorshidi, Samira
A2 Tita, George E.
A2 Carter, Jeremy
A2 Mohler, George
LA English
YR 2021
UL https://krimdok.uni-tuebingen.de/Record/1760914282
AB Crime diversity is a measure of the variety of criminal offenses in a local environment, similar to ecological diversity. While crime diversity distributions have been explained via neutral models, to date the environmental and social mechanisms behind crime diversity have not been investigated. Building on recent work demonstrating that crime rates can be inferred from street level imagery with neural network computer vision models, in this paper we consider the task of inferring crime diversity through street level imagery.
K1 Computervision
K1 Crime diversity
K1 Google Street View
K1 Geographically weighted regression
K1 Computer vision
DO 10.1007/s10940-021-09500-1