NUS Urban Analytics Lab

🥁 🥁 🥁 We are introducing Global Streetscapes, a large open labelled street view imagery dataset for urban science and analytics. It includes 10 million images each annotated with 346 attributes, accompanied by code and documentation we released openly as well. We hope you enjoy it and make use of it! A paper about the project, authored by Yujun Hou, Matias Quintana, Maxim Khomiakov, Winston Yap, Jiani Ouyang, Koichi Ito, Zeyu Wang, Tianhong Zhao, and Filip Biljecki, was published in the ISPRS Journal of Photogrammetry and Remote Sensing. Please read more on the website of the project: https://lnkd.in/g34NBzkQ The paper is available here: https://lnkd.in/gS-DYJdu (and here for free: https://lnkd.in/gnVUSXrH)

I attended IATBR 2024 in Vienna, Austria this week and presented my work on the causal impacts of visual elements on cyclist activities!  I met many great travel behavior researchers and got inspired to explore this field even more! The city and the University of Vienna were beautiful and very cyclist-friendly! Thank you Prateek Bansal for working with me on this project! It was also nice to connect with Behavioural Cognitive Science (BeCoS) Lab members: Maksimenko Vladimir, Jiaxuan Ding, Xinwei LI, and Zhuhan Jin! I am grateful to the NUS Department of Architecture and Filip Biljecki, NUS Urban Analytics Lab for funding this opportunity.

We're excited to continue our Virtual Summer School today with lectures from Angel Hsu, PhD and Filip Biljecki on AI for Buildings and Cities! Register for the program below to watch live: https://lnkd.in/dxdF9waH

We're excited to release Global Streetscapes - a massive open dataset of 10 million street-level images labelled with 300+ characteristics across 688 cities, designed to support urban science and analytics! 🛣️ 🌏 The project was led by Yujun Hou, in collaboration with Matias Quintana, Maxim Khomiakov, Winston Yap, Jiani Ouyang, Koichi Ito, Zeyu Wang and Tianhong Zhao, and conducted at our NUS Urban Analytics Lab. I am very proud of the team! The dataset was launched this week in tandem with its namesake article published in the ISPRS Journal of Photogrammetry and Remote Sensing, which details the motivation, related work, methodology, use cases, examples, and limitations. In a nutshell, this largest annotated dataset of street view imagery has been created to mitigate various challenges when using street view imagery. While this data has been absolutely useful, it lacks contextual and other information. The project builds upon contributions of thousands of Mapillary and KartaView contributors, contextualising millions of images with hundreds of attributes, taking their efforts forward for better urban analytics. In total, our dataset contains more than 3 billion records. We look forward to seeing the uses of the dataset and welcome collaborations! Read more about our project on its website: https://lnkd.in/gMvKyjky Link to the paper: https://lnkd.in/gJ4ZmfSa Free access to the paper in the next few weeks: https://lnkd.in/gM-ZC9N9 Many thanks to Yujun Hou for driving this effort. Follow her to stay tuned for further advancements. NUS Urban Analytics Lab, NUS Department of Architecture, NUS College of Design & Engineering

We are having an exciting and busy July! We are pleased to have hosted four faculty from overseas for visits and guest lectures: Adiwan Aritenang (ITB 🇮🇩), Piyushimita (Vonu) Thakuriah (Rutgers 🇺🇸), Ate Poorthuis (KUL 🇧🇪), and Fábio Duarte (MIT 🇺🇸). Big thanks to our guests for their time and the wonderful talks! Read more: https://lnkd.in/gvPaFShj NUS Department of Architecture, NUS College of Design & Engineering

Last week, I attended the 19th 3D GeoInfo Conference in Vigo, Spain, organised by the University of Vigo. At the conference, we presented our recent work — integrating human perception in 3D city models and urban digital twins, co-authored by Xiucheng Liang and Filip Biljecki. This work, for the first time, introduces an idea and method to integrate human perception of urban buildings into 3D city models, suggesting the potential for downstream studies. The conference paper is available here: https://lnkd.in/giDVF5nh Last year, our work 'Humans as Sensors in Urban Digital Twins' took the first step to develop a conceptual framework and integrate the perception of thermal comfort in urban digital twins. This year, we focused on visual perception, supporting better city planning and human-centric digital twins. This research is part of my PhD, contributing to a shift towards human-centric urban digital twins from a socio-technical perspective. We believe that the new subjective attributes incorporated into existing 3D buildings open the door for new applications. Many thanks to the conference committee Lucía Díaz Vilariño and others for organising such a wonderful event. I had the opportunity to reconnect with our community and engage with new researchers in Vigo. Very looking forward to the next 3D GeoInfo Conference in Tokyo, Japan, in 2025!

Did you know? Aichi Prefecture in Japan boasts its own Merlion and is the original home to Singapore's first Japanese resident, Yamamoto Otokichi. Otokichi played a pivotal role in Japan's foreign trade and settled in Singapore in 1862. Today, Aichi Prefecture is known as Japan’s industrial heartland. Its capital city, Nagoya, is renowned not only for Ghibli Park but also serves as a global manufacturing hub, hosting iconic brands like Toyota, Rinnai, Brother, Noritake, and Mitsubishi Heavy Industries. Nagoya has a rich historical legacy, being the birthplace of Japan's first shogun, Minamoto Yoritomo, and pivotal figures like Oda Nobunaga, Toyotomi Hideyoshi, and Tokugawa Ieyasu. These figures later established the capital cities of Kyoto, Osaka, and Edo (now Tokyo), laying the foundational roots of modern Japan. Locally, there's a good-natured quip that Nagoya built Tokyo and Osaka. Recently, I was privileged to be part of the inaugural PhD cohort at NUS Overseas Colleges, immersing myself in Nagoya's dynamic culture. During my six-months internship at MAP IV, Inc., a self-driving car startup from Nagoya University, I collaborated with talented engineers (Yuki Kitsukawa, Jacob Lambert, Patiphon Narksri, Abraham Monrroy Cano) and explored how deep learning can be used on citywide 3D data to improve navigation of self-driving cars in complex urban environments. A highlight was engaging with stakeholders from government, industry, and academia on urban challenges. I represented Singapore at a round table with Japanese Minister of Education, Minister Moriyama and Nagoya University President, Prof. Sugiyama on international student experiences in Japan. I also met with Mr. Numasawa, Director-General of Policy Planning at the Aichi Prefectural Government, and learned about startup initiatives shared between Nagoya University and NUS. Visiting Takenaka Corporation’s Design Division (Hajime Iseda, Genki Unno), I was heartened to discover initiatives in net-zero buildings, circularity, and spatial infrastructure management for sustainability. I also had the pleasure of meeting with President of Prodrone (Tony Shunsuke TOYA), where I got to learn about innovative drone applications. At Nagoya University, I was inspired by the work at Prof. Nobuo Kawaguchi's Lab on real-world human mobility and activity datasets. I gained additional insight on the use of computational tools to support diverse urban use cases such as combating urban heat island effects and supporting urban development at Prof. Takashi Asawa's and Prof. Takuya Oki's Lab at the Tokyo Institute of Technology. My time in Japan broadened my worldview, made possible by the NOC team’s (Tun Leng LEE, Gean C.) meticulous planning, and the warm hospitality of Prof. Hiroshi Imaizumi and Kazuya Takeda. Special thanks to my PhD supervisor, Prof. Filip Biljecki, whose invaluable guidance and support enabled me to leverage this exchange for my research and career goals effectively.

On Wednesday 17 July, we are happy to host Fábio Duarte from the MIT Senseable City Lab. He will give a talk titled What happens with big data approaches to urban science when data don’t exist? Join us at our NUS Department of Architecture & NUS College of Design & Engineering next week. More about our seminars: https://ual.sg/seminars/

On Tuesday 16 July, we are glad to welcome Ate Poorthuis, Asst Prof at KU Leuven - Department of Earth and Environmental Sciences. He will give a talk titled Twenty years of using crowd-sourced geodata to analyze urban phenomena – where do we go next? Join us at our NUS Department of Architecture & NUS College of Design & Engineering next week. More about our seminars: https://ual.sg/seminars/