Matt Jacobsen and Pingfan Meng’s research on object tracking won the Best Poster Award for Computer Science and Engineering at the Jacobs School Research Expo on Thursday April 17. From the Jacobs School Press release: They developed a computer vision tracking system that is faster and more accurate than the current state of the art. They did so by devising an algorithm that divides the processing between software running on a CPU and custom hardware implemented using a field-programmable gate array. This system can track a single target at 1160 frames per second or 57 independent targets at 30 frames per second. That’s 68 times faster than an approach that uses software only. This considerable increase in computing power improves the accuracy of the tracking algorithm by tracking multiple targets on the object under consideration. For example, instead of tracking a hand using a single target, the system will track six objects – each of the fingers and the palm – making it significantly more accurate.