Hamed Tabkhi
University of North Carolina at Charlotte, NC·Computer Science
Security cameras are everywhere in cities, but using their video raises a hard problem: how can it make public spaces safer and better designed without tracking who people are? This research builds artificial intelligence (AI) that studies how people move and behave instead of who they are.
- Fields
Computer Science
Engineering
- Topics
computer vision
deep learning
artificial intelligence
- Grants
- 3 known grants, through Sep 2027
What they work on
- Develop a semi-supervised transformer model that detects unusual or threatening behavior in real time from surveillance video. Use identity-neutral visual features so that no personally identifiable information is needed.
- Create a new anomaly detection video dataset and realistic metrics for detection reliability. Test the system's reliability in real-world pilot deployments.
- Convert raw camera pixels into anonymous motion tokens and heatmaps on small computers at each camera (edge computing). Detect walking paths, dwell times, queue lengths and unusual behavior in downtown Charlotte.
- Build an online active-learning framework that tunes detection settings using feedback from security staff and site managers. Feed results into dashboards, crowd-density maps and occupancy forecasts.
- Assess demand for a privacy-preserving urban video analytics platform through interviews with urban planners, transportation agencies and business districts. Ask how it could measure the impact of infrastructure changes.
More about the research
It turns ordinary camera feeds into anonymous data that can flag safety threats and guide urban planning. The projects share one core technology and differ in purpose: threat detection, a downtown pilot deployment, and testing whether the platform could become a product.
Why it matters
Cities, businesses and transit agencies could gain timely data on how people use sidewalks, parks and downtown spaces without storing faces or other personal information. This could guide decisions about lighting, signage, staffing and infrastructure spending. Behavior-based threat detection with existing cameras may also offer lower-cost, less biased security for public agencies, businesses and critical infrastructure.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
CIVIC-FA Track B: Strengthening Community Safety and Well-being with Ethical, AI-Assisted Video Solutions
NSF · Oct 2025 – Sep 2027
PFI-TT: Behavioral Analysis for Safer Communities: Fair and Ethical AI for Trusted Surveillance
NSF · Aug 2023 – Jul 2027
I-Corps: Translation Potential of Artificial Intelligence-Assisted, Data-Informed Urban Planning
NSF · Jun 2025 – May 2027
Closest labs
David Anastasiu
Santa Clara University, CA·Computer Science·Early-career
Create a digital-twin framework that semi-automatically captures and annotates video anomaly anticipation datasets from static camera viewpoints, recomposing real objects and motion patterns into new motion anomalies.
Yan Yan
University of Illinois Chicago, IL·Computer Science
Develop unsupervised learning models for re-identifying and tracking people and vehicles across a network of city cameras, processing video in real time across edge devices and the cloud.
Assefaw Gebremedhin
Washington State University, WA·Computer Science
Develop robust and trustworthy machine learning algorithms with anomaly-detection features that flag malicious activity in AI systems used in healthcare, autonomous vehicles and smart cities.
The school
University of North Carolina at Charlotte: 9 labs on PI Pages, $95M a year on research, 196 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.