The MLAI Meetup is a community for AI researchers and professionals which hosts monthly talks on exciting research. Our format is:
- 6:00 - 6:20: Socializing
- 6:20 - 6:40: Announcements and AI news
- 6:40 - 7:40: Talk(s) and Q&A
- 7:40 - 8:00 Networking
- 8:00: Head to the nearest pub for dinner
Shinjita Das: When Maps Aren’t Enough: Using Spatial AI to Measure the Green We See
Talk description: A tree can cool a street, appear clearly on a map – and still be completely invisible from someone’s window. That gap between what exists in a city and what people actually see and experience has shaped much of my research journey. When I first started working with geospatial science, I was fascinated by the idea that spatial data could reveal patterns in cities that are difficult to notice on the ground. During my Master’s research, I used deep learning to detect individual urban trees and examined their contributions to urban cooling. That was my first real encounter with machine learning, and it changed the way I thought about maps: they were no longer just tools for showing where things are, but platforms for asking what else we could learn from the environment.
That curiosity led me further into GeoAI. If machine learning could help identify trees and quantify their environmental role, could similar computational approaches help us understand how people actually experience greenery? This question followed me into my PhD, where my toolkit expanded from 2D GIS and spatial statistics to 3D city modelling, computer vision and emerging GeoAI approaches. At the same time, the research question became more human. Instead of simply asking “Where is the green?”, I began asking, “Who can actually see it and how does the view impact mental health?”
In this talk, I’ll trace that journey through examples from my research: from detecting trees and measuring their cooling effects to reconstructing urban environments in 3D and modelling greenery from the perspective of apartment residents. Along the way, I’ll explore how GeoAI, computer vision and spatial modelling can help us move beyond mapping cities from above and start understanding them from the perspective of the people who live within them.
Speaker bio: Shinjita Das is a PhD candidate in the School of Global, Urban and Social Studies and the Department of Mathematical and Geospatial Sciences at RMIT University. Her ARC-funded High Life Study research examines apartment living, urban greenery and mental health using 3D GIS modelling and spatial statistics. Her interest in machine learning began during her Master’s research, where she used Faster R-CNN to detect urban trees and investigate their cooling effects. As a member of RMIT’s GISail research group, her broader research interests lie in GeoAI, spatial machine learning, 3D urban modelling, urban heat vulnerability, environmental exposure assessment, health-integrated spatial analysis, and the use of emerging geospatial technologies to support healthier, more sustainable, and liveable cities. She is also a tutor and aims to cultivate critical thinking and innovation in spatial science, helping foster the next generation of geospatial professionals equipped to address complex urban and environmental challenges.
