Deep learning

Human behaviour on networked systems

Networked systems went through several phases over the past 40 years. In the first phase the network was a vessel for movement of data. The second processed and stored that data in order to allow users access to services that provide unique value like music, entertainment, and e-commerce, and the third was primaily powered by large companies that added social layers to these troves of services thereby allowing effective curation and recommendation of products and ads.

Quantifying the Urban Exposome

The 21st century would witness one of the largest urban migration in human history. A lot of research has gone into understanding the effects of the urban life on physical and mental health. The results have been unanimous: The urban environment has long lasting impacts on our well being. The facets of our urban environment that impact on our health and well being are called the urban exposome. In these projects, I attempt to quantify this exposome using techiques from deep learning, complex network, GIS, and NLP.

GraphNLI: A Graph-based Natural Language Inference Model for Polarity Prediction in Online Debates

Online forums that allow participatory engagement between users have been transformative for public discussion of important issues. However, debates on such forums can sometimes escalate into full blown exchanges of hate or misinformation. An …

Imagine a Walkable City: Physical activity and urban imageability across 19 major cities

Can the shape of a city promote physical activity? The question of why individuals engage in physical activity has been widely researched, but that research has predominantly focused on socio-demographic characteristics (e.g., age, gender, economic …

How epidemic psychology works on Twitter: evolution of responses to the COVID-19 pandemic in the US

Disruptions resulting from an epidemic might often appear to amount to chaos but, in reality, can be understood in a systematic way through the lens of “epidemic psychology”. According to Philip Strong, the founder of the sociological study of …

The Healthy States of America: Creating a Health Taxonomy with Social Media

Since the uptake of social media, researchers have mined online discussions to track the outbreak and evolution of specific diseases or chronic conditions such as influenza or depression. To broaden the set of diseases under study, we developed a …

Jane Jacobs in the Sky: Predicting Urban Vitality with Open Satellite Data

The presence of people in an urban area throughout the day -- often called 'urban vitality' -- is one of the qualities world-class cities aspire to the most, yet it is one of the hardest to achieve. Back in the 1970s, Jane Jacobs theorized urban …

FaceLift: a transparent deep learning framework to beautify urban scenes

In the area of computer vision, deep learning techniques haverecently been used to predict whether urban scenes are likely tobe considered beautiful: it turns out that these techniques areable to make accurate predictions. Yet they fall short when …

Facelift makes it to top 16 Ph.D. projects across the U.K. @ EPSRC Pioneers Competition

Our project FaceLift: A deeplearning pipeline to beautify urban spaces, has been shortlisted in the highly selective EPSRC Pioneers competetion. We are one of the top 16 Ph.D. projects around U.K. It stands in the top 4 projects among the creative computing category. Please find the full announcement here.

Mapping and Visualizing Deep-Learning Urban Beautification

Information visualization has great potential to make sense of the increasing amount of data generated by complex machine-learning algorithms. We design a set of visualizations for a new deep-learning algorithm called FaceLift …