Visual Analysis and Synthesis with Physically Grounded Constraints

Tuesday, March 15 at 10 am in the Games and Visual Narrative classroom, Venture 2 Suite 530.

Title: Visual Analysis and Synthesis with Physically Grounded Constraints

Abstract:
“The past decade has witnessed remarkable progress in image-based, data-driven vision and graphics. However, existing approaches often treat the images as pure 2D signals and not as a 2D projection of the physical 3D world. As a result, a lot of training examples are required to cover sufficiently diverse appearances and inevitably suffer from limited generalization capability.

In this talk, I will present “”inference-by-composition”” approaches to overcome these limitations by modeling and interpreting visual signals in terms of physical surface, object, and scene. I will show how we can incorporate physically grounded constraints in a non-parametric optimization framework. I will use two specific examples in computational photography to demonstrate the benefits of these constraints: (1) revealing the missing parts of an image due to the removal of a foreground or background element and (2) recovering high spatial frequency details that are not resolvable in low-resolution observations. The resulting algorithms are simple and intuitive while achieving state-of-the-art performance without the need of training on an exhaustive set of visual examples. I will end my talk with a brief discussion of broader applications of visual synthesis and analysis.”

Bio:

“Jia-Bin Huang is a Ph.D. candidate in the Department of Electrical and Computer Engineering at University of Illinois, Urbana-Champaign advised by Prof. Narendra Ahuja. His research interests include computer vision, computer graphics, and machine learning with a focus on visual analysis and synthesis with physically grounded constraints. His research received the best student paper award in IAPR International Conference on Pattern Recognition (ICPR) in 2012 for the work on computational modeling of visual saliency and the best paper award in the ACM Symposium on Eye Tracking Research and Applications (ETRA) in 2014 for work on learning-based eye gaze tracking. Huang is the recipient of the UIUC Graduate College Dissertation Completion Fellowship (2015), Thomas and Margaret Huang Award for Graduate Research (2015), Best Talk Award at CSL Student Conference (2016), Beckman Cognitive Science/Artificial Intelligence Award (2015), Sundaram Seshu Fellowship (2014), MOE Technologies Incubation Scholarship (2014), and the PURE Best Research Mentor Award (2012).

Personal website: http://www.jiabinhuang.com“