Hao Su: 2014 Networks of Shapes, Images, and Programs: The Power of Joint Data Analysis Workshop

 

Wednesday, April 16, 2014
Location: Fisher Conference Center, Arrillaga Alumni Center

"Exploiting Mixed Image and Shape Networks"
12:00pm - 12:15pm

Abstract:

Images are easy to acquire, view, publish, share and as a result ubiquitous. They, however, lack critical depth information. This poses a serious bottleneck for many image manipulation, editing, and retrieval tasks. In this talk we, consider the problem of 'lifting' an image of an object to 3D by exploiting a collection of aligned 3D models of similar or related objects. Our key insight is that even when the imaged object is not contained in the shape collection, the network of shapes implicitly characterizes a shape-specific deformation subspace that regularizes the problem and enables robust diffusion of depth information from the shape collection to the input image. We evaluate our fully automatic approach on diverse and challenging input images, and demonstrate applications to depth-enhanced image editing.


Bio:

Hao Su is currently a Ph.D student in the Computer Science Department of Stanford University, advised by Prof. Leonidas Guibas. His research interest is to understand the semantic and geometric information from visual medium such as images and scan data. To tackle this very challenging problem, his previous work involves crowd-sourcing, statistical learning, convex optimization, geometry processing and graph retrieval. Hao received his bachelor degree from the Department of Computer Science of Beihang University in July 2006. During 2005-2008, he was a Research Intern in Natural Language Computing Group and Visual Computing Group of Microsoft Research Asia, co-advised by Dr. Harry Shum, Prof. Wei Li, Dr. Jian Sun and Dr. Ming Zhou. During 2008-2010, he worked with Prof. Fei-Fei Li and Prof. Silvio Savarese in Princeton University. During 2010-2011, he worked with Prof. Fei-Fei Li and Prof. Andrew Ng in the AI lab of Stanford University.