Computer and information sciences professor Haibin Ling has been awarded the National Science Foundation Career Award of $479,691 for his work on bettering computer image analysis for a variety of uses in society.
Ling’s five-year project, “High-order Tensor Analysis for Groupwise Correspondence: Theory, Algorithms and Applications,” looked at a range of topics like computer communications and applications, image analysis and the early detection of diseases like osteoporosis.
According to the National Science Foundation’s website, Ling’s project “develops a unified framework for this problem – finding correspondence among a group of objects – and to apply the framework to different applications.”
“Basically, we try to make the computer understand what it says,” Ling said. “My project is mainly about fundamental tools to support computer applications.”
Ling said fundamental tools to support communication included cameras having the ability to localize eyes in a snapshot and zoom in.
“Our interest is to use this image analysis to understand medical images,” Ling said.
With this technology, it is possible to predict diseases such as osteoporosis based on medical and dental images taken in a doctor or dentist’s office. Dentistry professor Jie Yang collaborated with Ling and managed the project.
Ling’s project also deals with sending a warning message to Internet users before they upload questionable content online.
In 2010, a Rutgers University student committed suicide after his roommate and a hall mate used his webcam to secretly tape him kissing his partner and later uploaded it to the Internet. Ling said he believes technology can prevent incidents like this.
“In that time, I was thinking that when people publish such videos on websites, you can tell a difference between regular taped videos and secretly taped videos,” Ling said.
In regular videos, the lighting is typically good and the posing is natural, while in secretly taped videos the content is often strange, the lighting is dark and the posing seems odd.
Ling said it is still a work in progress, but he hopes to be able to use image analysis to identify potentially harmful images and send out a warning message before uploading to prevent future tragedies.
The award money given to Ling will grant him one month’s summer salary per year, as well as funding the education of a Ph.D. student throughout his five-year education. Additionally, the money will apply associated funding for potentially two to three undergraduate students involved in research.
Logan Beck can be reached at firstname.lastname@example.org.