Progress in contemporary visual information processing is driven by the joint pursuit of interactivity, realism, and flexibility. To find viable solutions to the scientific and technological challenges encountered towards these goals, we pursue interdisciplinary strategies that draw on computer graphics, computer vision, applied optics, and human visual perception. Our goal is to contribute answers to fundamental scientific questions that enable us to devise new algorithms for visual computing applications.
Our work on Comparative analysis of three different modalities for perception of artifacts in videos has been accepted to ACM Transactions on Applied Perception and will be presented at ACM Syposium on Applied Perception.
Optical Flow-based 3D Human Motion Estimation from Monocular Video