A Professor from the University at Albany, Siwei Lyu, has won the Faculty Early Career Development (CAREER) Award from the National Science Foundation (NSF). The prize comes with a $500,000 award spread over five years, which Lyu will use to develop “A New Statistical Framework for Natural Images with Applications in Vision”, a project that will help detected altered digital images.
The applications could be far reaching, and Lyu is currently pursuing collaborations with the New York State Police Forensic Investigation Unit to use his digital image techniques in a criminal investigation setting. The process involves using new statistical models to analyze the images. With tools like Photoshop available to everyone from a high school students looking to make a joke image, to professionals looking to forge a document, the authentication tools Lyu hopes to introduce couls have legal, commercial, and practical settings.
Lyu’s CAREER project will also help to better understand biological vision, create new tools for faster and cleaner image compression transmissions, restore corrupted images, and automatically detect and recognize objects in an image.
“The wisdom of the past two decades has been that images are formed by adding and subtracting simple units. For instance, by adding up images of trees, you should get an image of a forest,” Lyu said. “But this is not the way images from our physical world are created. You put images of trees on top of each other based on how close they are from you to get an image of a forest, and this nonlinear process is very different from simply adding them together.”
More information on Lyu and his project can be found on his personal webpage.
- Microsoft unveils $3,500 HoloLens 2 at MWC 2019. Here’s what you need to know
- The best movies on Netflix right now (March 2019)
- 11 Hollywood heavyweights who jumped from the big screen to the small screen
- The best movies on Amazon Prime right now (March 2019)
- Learn something new, from coding to marketing, with these free online courses