Public Release: 

Computer science researcher teams with industry to provide facial recognition technology

Computer vision

University of Texas at Arlington


IMAGE: Vassilis Athitsos, an associate professor in the UTA Computer Science and Engineering Department, is working with a global firm to evaluate existing deep learning methods for face detection and facial... view more 

Credit: UT Arlington

From security to self-driving cars, computer vision is becoming an important part of society, and the information technology industry is ramping up efforts in the area to capitalize on emerging opportunities.

Vassilis Athitsos, an associate professor in the Computer Science and Engineering Department at The University of Texas at Arlington, is working with a $59,463 grant from Macnica Americas to evaluate existing deep learning methods for face detection and facial recognition to determine how the company could improve performance or reduce computational load and make better technology available to the company's customers.

Macnica Americas is a fully franchised semiconductor distributor covering North America. It is a division of Macnica Inc., a $3.2 billion global leader in semiconductor distribution and design services.

Computer vision, or facial recognition, tracks where humans are and how they move. It has many applications, such as unlocking a laptop computer or cell phone, or controlling who can or cannot enter a building. On a larger scale, facial recognition could be used for surveillance of people on terrorist watch lists, but to do so would require scaling up the software to quickly and accurately recognize one or two faces among thousands in an airport or on a city street, and that technology does not yet exist.

Athitsos, doctoral student Amir Ghaderi and master's student Saif Sayed agreed to work with Macnica because they saw potential to approach their research from a different point of view where research results in a product going to market and not only a paper in an academic journal.

"Ten years ago, computer vision wasn't relevant to the economy, but now companies are investing a lot in the technology," Athitsos said. "It's getting to the point where it's worth asking how or if our work is commercially deployable, and if not, how can we make it so?"

Athitsos said working with Macnica has given the team a new point of view.

"We're doing something useful for them, helping them to transfer technology to the marketplace, and they're helping us to refine our thinking outside of the scope of academia," Athitsos said.

Ghaderi and Sayed also are benefitting from the collaboration. Sayed is interning at the company and Ghaderi has taken on research for the project at UTA, which will allow him to formulate his own research questions and begin writing papers to add to existing knowledge.

"All of the work that Amir and Saif are doing on this will be highly relevant and make them more marketable as they seek jobs after earning their degrees," Athitsos said.

This research is just one example of innovative thinking from the Computer Science and Engineering Department in the area of data-driven discovery, one of the themes of UTA's Strategic Plan 2020: Bold Solutions | Global Impact, said Hong Jiang, Wendell Nedderman Chair and Professor in the department.

Other examples include Athitsos' research that used computer vision to study how a computer can recognize signs in American Sign Language and enable users to search dictionaries of American Sign Language to look up the meaning of an unknown sign, and a current grant by Fillia Makedon, a computer science and engineering professor, who is using computer vision and machine learning to help experts assess learning difficulties in children.

"The heart of computer science and engineering is finding ways to use computers more efficiently to improve our interactions with the world around us," Jiang said. "This partnership is an example of how knowledge developed in an academic setting can be applied in industry with benefits for all involved."


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.