Making a New City Image…or, an Eye for AI

Making a New City Image explores the machine-mediated perception of urban form: new ways of seeing, understanding and experiencing cities in the information age.

Student
Brian Ho (MDE ’18)

Project Type
Independent Design Engineering Project

Year
Fall 2017 – Spring 2018

Project Advisors
Krzysztof Gajos
Robert Pietrusko

Spanning the disciplines of urbanism and computer science, this project fosters a productive dialogue between the two. What might we learn about the city using computer vision, deep learning, and data science? How might the history, theory, and practice of urbanism — which has long viewed the city as a subject of measurement — inform modern applications of computation to cities? Most importantly, can we ensure that both disciplines understand the city as it is perceived by people?

This project also formalizes a practice of computer vision cartography. First, it has revisited Kevin Lynch’s Image of the City, applying machine learning models to archival photographs and historical maps of Boston, and classifying from them Lynch’s five elements of the city image. This process has also been adapted to the present day. I have created instruments and devices for the procedural capture of street-level imagery (on bike or by foot), and the automated identification of new categories of urban environments crowdsourced from human input.

Together, these methods produce a new mode of analysis that balances a comprehensive perspective at the scale of the city with a focus on the texture, color, and details of urban life.

Process 1: Inputs

How can urban theory and history inform the data inputs for computational models of cities?

By training a convolutional neural net to classify the elements of the city image from geocoded and labeled images from the Kevin Lynch archive (accessed via Flickr API), we can combine historical and theoretical perspective with new digital approaches.

Images from the Kevin Lynch archive were geocoded and labeled using a re-projection (created in GIS) of his original city image diagram. The resulting geocoded images can be viewed online.

A white city grid on a black background with blue dots scattered across it. Text on the left side of the image reads: "The Image of the City, redux. This book is about the look of cities, and whether this look is of any importancee, and whether it can be changed. The urban landscape, among its many roles, is also something to be seen, to be remembered, and to delight in. Giving visual form to the city is a special kind of design problem, and a rather new one at that. Kevin Lynch, in The Image of the City".

Process 2: Instrument

Eye for AI cameraHow can new digital methods extend the knowledge of urbanism and design practice?

By building a device to capture 360-degree imagery, associated to a GPS tag, we can build our own dataset to represent the city of Boston today.

A white on blue schematic of a camera mounted on bike handlebars meant to capture the city image.

An example of an image captured with a 3D camera for the project, showing a wide-angle view of a Boston-area intersection. Brian's arms are each visible in the left and right sides of the frame.

Process 3: Interface

How can digital methods capture human perspective, to inform both urbanism and computation?

By developing an interface for crowdsourced labeling, we can collect human perception at scale. These label can be used to inform machine learning models, and can voluntarily be attributed to their contributors.

Learn more