Researchers Develop Real-time AI Testbed

Dr. Hamed Tabkhivayghan

Research Highlight

Hamed Tabkhi and his team have extensively engaged and partnered with Central Piedmont Community College (CPCC) to create the first community-based real-time Artificial Intelligence (AI) testbed at the CPCC campus. The effort is supported by a $1.9M research grant from the National Science Foundation, Smart and Connected Community Program (award# 1831795). The testbed has been used as a real-world demonstrator of real-time AI for privacy-first intelligent processing and services. The students both at UNC Charlotte and CPCC play an active role in developing the technology, installing the testbed, and evaluating which includes creating the labeled dataset for the training of AI algorithms. In collaboration with the CPCC Criminal Justice program, the team is set to expand the testbed to a campus-wide coverage across many cameras to explore and validate the practicality of developed algorithms in real-world scenarios with a privacy-first design mindset.

A real-time demo of the technology at the CPCC campus is presented here. The demo demonstrates real-time pedestrian detection, pose estimation, action detection, and scene segmentation at the edge without performing any facial recognition or extracting Personally Identifiable Information (PII). The success of this project has stemmed from the unique collaboration between faculties in the computer engineering discipline and criminal justice and criminology. The other principal members of the team are Dr. Shannon Reid, associate professor of Criminal Justice and Criminology, and Dr. Arun Ravindran, associate professor of Electrical and Computer Engineering from UNC Charlotte, and Ms. Jeri M. Guido the Program Chair of Criminal Justice Technology at CPCC.