Deep Learning the City : Quantifying Urban Perception At A Global Scale

TL;DR

This paper presents a large-scale crowdsourced dataset and a neural network model to quantify urban perception across the globe, enabling scalable analysis of city environments and their impact on residents.

Contribution

It introduces a new extensive dataset of urban images and pairwise comparisons, and develops a neural network model to predict human perceptions of cities at a global scale.

Findings

Crowdsourcing combined with neural networks can effectively quantify urban perception.

The dataset includes over 110,000 images and 1.17 million pairwise comparisons.

The model accurately predicts human judgments of urban attributes.

Abstract

Computer vision methods that quantify the perception of urban environment are increasingly being used to study the relationship between a city's physical appearance and the behavior and health of its residents. Yet, the throughput of current methods is too limited to quantify the perception of cities across the world. To tackle this challenge, we introduce a new crowdsourced dataset containing 110,988 images from 56 cities, and 1,170,000 pairwise comparisons provided by 81,630 online volunteers along six perceptual attributes: safe, lively, boring, wealthy, depressing, and beautiful. Using this data, we train a Siamese-like convolutional neural architecture, which learns from a joint classification and ranking loss, to predict human judgments of pairwise image comparisons. Our results show that crowdsourcing combined with neural networks can produce urban perception data at the global scale.