How are urban design qualities associated with perceived walkability? An AI approach using street view images and deep learning

Abstract

This study attempted to provide further understanding of the association between walkability and urban design qualities in Xi’an, China. The novelty of this paper is to employ street view image data to predict people’s perceptions of walkability by constructing deep convolutional neural networks. In general, our results confirm the relationship between urban design qualities and perceived walkability in a mixed area of modern and traditional cityscapes and urban forms. Specifically, the most dominant variable is complexity, which has both positive and negative impacts, and imageability and transparency are negatively linked to walkability. It was also found that the impact of five urban design qualities on walking perception is clearly spatially differentiated.