Prof. JIN Ying
Director of the Martin Centre for Architectural and Urban Studies, Professor of Architecture and Urbanism, Deputy Head (Research), Department of Architecture, University of Cambridge
Biography
Professor Jin is the current Director of the Martin Centre for Architectural and Urban Studies, Deputy Head (Research) of the Department of Architecture at University of Cambridge.
His main research interests are computer models of cities, and urban history. He has extensive industry experience and directed multi-disciplinary teams in building and using computer models as experimental platforms to appraise policy scenarios that involve investment, regulation, pricing and promotional campaigns. Key projects include strategic planning of London and surrounding regions, sub-regional and local planning in the English Midlands, transport and energy scenarios for the European Union, long term city region and transport plans in China and in South America, mapping urban poverty in emerging economies, and assessing development and transport options for Cambridge and surrounding regions. His interests in urban history lie mainly with the European Renaissance cities and the Chinese cities since the Tang Dynasty in the 7th Century.
At the Department of Architecture Professor Jin leads the Cities and Transport Research Group, which is one of the world’s leading centres in the creation and use of conceptual and practical models for cities and city-regions. These models have been applied in policy and planning studies to assess novel designs of buildings, neighbourhoods, transport and energy systems. The group’s past policy impacts were reviewed in a Cambridge University case study in REF2014.
Topic
The Role of Deep Learning for Understanding and Designing City Streets
Abstract
This lecture considers the future of spatial data analysis and data-driven urban design in the age of AI tools including ChatGPT. How will the roles change in this field for urban designers, the wider public and government as the AI tools become available? The lecture uses the example of street design – one of the fundamental aspects of cities – to show how we are able to (1) harness deep learning (which underlies much of the current AI tool box) in understanding the streets and neighbourhoods of a diverse range of cities in Europe, and (2) create new design tools for preserving the identities of local street fabric while allowing expert and lay users the freedom to adapt to new needs in the cities. Through the PhD work of Dr Fang Zhou at University of Cambridge the deep learning tools have been applied and tested in Siena, Perugia, Rome and Florence for physical terrain effects and in Amsterdam, Barcelona, Berlin and Prague for historic and cultural influences.
