Dense Street Imagery (DSI)
Visual data from digital eyes on the street.
Under Review
Project Team
Matt Franchi
Computer Science PhD Candidate
Hauke Sandhaus
Information Science PhD Candidate
Madiha Zahrah Choksi
Information Science PhD Candidate
Severin Engelmann
Digital Life Initiative Postdoctoral Associate
Wendy Ju
Associate Professor
Helen Nissenbaum
Professor / Director of Digital Life Initiative
What is dense street imagery?
Dense Street Imagery (DSI) represents a breakthrough in visual urban data collection, combining advanced vehicle hardware, imaging technology, and networking capabilities to create dynamic, real-time depictions of city environments. Unlike traditional static snapshots from services like Google Street View, DSI leverages temporal density through networked dashcams and driver-assist systems to deliver fresh, continuous imagery at unprecedented frequency.
The Scale of Digital Eyes
Existing Applications
Weather Events
Tracking transient environmental conditions in real-time across urban areas
Infrastructure Monitoring
Documenting changes in urban infrastructure like scaffolding and construction
Traffic Patterns
Analyzing vehicle placement and flow to improve urban mobility
Road Conditions
Real-time assessment of physical road quality and maintenance needs
DSI vs Traditional Street Imagery
Traditional Street View
- Updated every few years
- 10 million miles of footage
- Fixed camera positions
Dense Street Imagery
- Continuous, real-time updates
- 200+ million miles of driving imagery
- Distributed mobile camera network
Our research explores both the potential applications and privacy implications of this emerging data source. Using a dataset of 8 million dashcam images captured throughout New York City, we're examining how these "digital eyes on the street" might reshape our understanding of urban spaces while addressing critical questions about privacy and contextual integrity.