Software mimics humans with image matching
Researchers at Carnegie Mellon University's School of Computer Science have developed software that can enable a computer to mimic the human ability to find visually similar images.
Researchers atCarnegie Mellon University's School of Computer Science have developed software that can enable a computer to mimic the human ability to find visually similar images.
The research team, led by Alexei Efros, associate professor of computer science androbotics, and Abhinav Gupta, assistant research professor of robotics, found that their technique performed well on a number of visual tasks that normally stump computers, including matching sketches of automobiles with photographs of cars.
In addition to automated image searches, the new technique has applications in computational rephotography: the combination of historic photographs with modern-day photos taken from the same perspective.
Future work includes using the technique to enhanceobject detection for computer vision and investigating ways to speed up the computationally intensive matching process.
Aresearch paper outlining the technique is available.
-- By Dave Wilson, Senior Editor,Vision Systems Design