 |
Center for Nonlinear Studies |
Goksel Dedeoglu is a PhD candidate in the Robotics Institute at Carnegie Mellon University. He received a B.S. in Control and Computer Engineering from Istanbul Technical University in 1997, and an M.S. in Computer Science from the University of Southern California in 2000. His current research focuses on enhancing low-resolution face videos by means of space-time models and priors. "Face Hallucination" aims to recover high quality, high resolution images of human faces from low-resolution, blurred, and degraded images or video. We approach this ill-posed problem with domain-specific models and priors that can generate plausible high-frequency image details. In this talk, I will present some of our recent formulations and algorithms. I will first introduce a data-driven approach with a non-parametric video model. The generative model is a 3-D Markov Random Field (in space and time), where nodes correspond to local video patches, and links encode spatio-temporal consistencies as well as image formation & degradation processes. Hallucination is posed as a probabilistic inference problem, seeking the high resolution video composition that best satisfies the constraints expressed through the graphical model. In the second part, I will present a parametric model-fitting option. For this we adopt the Active Appearance Model (AAM), commonly used in visual face tracking and interpretation tasks. After exposing the shortcomings of the traditional AAM fitting approach in low-resolution scenarios, we develop a "resolution-aware" fitting criterion. By explicitly accounting for the finite size sensing elements of digital cameras, our fitting method simultaneously models the processes of object appearance variation, geometric deformation, and image formation. Experimental results confirm the viability of face hallucination using AAMs as well. Joint work with Takeo Kanade, Simon Baker, and Jonas August.
Host: DDMA Speaker Series