Vishal Monga | Electrical Engineering
Humans and machines alike are often engaged in visual tasks that involve analyzing, interacting with, forming or improving image data. The application landscape of automating visual tasks is rich and diverse and spans consumer and medical imaging, robotics and vision, remote sensing and space sciences, smart systems such as those for traffic analysis and process control. In the past decade, machine learning algorithms have greatly accelerated advances in automating visual tasks -- yet they are overly dependent on the quantity and quality of training image data available. This talk will survey research topics pursued in the Information Processing and Algorithms (iPAL) lab at Penn State (http://signal.ee.psu.edu), which is focused on developing innovations for visual tasks such as image quality enhancement, segmentation and classification when available data is severely limited, noisy or non-ideal.