MASC Research Areas


Recent Highlights


Material Transfer
image
Guilin Liu and researchers at Adobe created a Physically Based Rendering Network that can synthesize realistic re-rendering of a single object with different surface materials. (ICCV, 2017 Fall)

Layer Decomposition
image
We proposed a way to decompose a single image into layers of monochrome image via RGB-space geometry. (TOG, 2017 Summer)


image
Disjoint Convex Shell
We developed a software tool that turns a shape into a set of disjoint convex objects and made paper crafting much easier. See the project page: DC shell for detail. (SPM, 2017 Spring)



Making Shadow Art
Making Shadow Art
Our work on shadow art is featured on the cover page of magazine Hyperseeing. (2016 Summer)







image
Local Minkowski Sum
We construct local Minkowski sum to enable continuous penetration depth on complex shapes. (2016 Spring)




Learning to Segment and Unfold
image We proposed to simultaneously segment and unfold a non-convex mesh into foldable patterns by learning from failed unfoldings. This makes paper crafting easier. (2016 Spring)



image
Continuous Visibility Feature
A new type of visibility measurement named Continuous Visibility Feature (CVF) is proposed. CVF better encodes the surface and part information of mesh than the tradition line-of-sight based visibility. CVF can be used for many shape analysis tasks. (2015 Spring)

image
Convex Ridge Separation
We investigate an approach that decomposes a mesh based on the identification of Convex Ridges. Intuitively, convex ridges are the protruding parts of the mesh. (2014 Fall)


Dual-Space Decomposition
image We proposed a new decomposition method, called Dual-space Decomposition that handles complex 2D shapes by recognizing the importance of the negative areas (e.g. holes). This work is to be published in CVPR 2014. (2014 Spring)

image
Folding rigid origami
We developed an adaptive randomized search to fold rigid origami. Ph.D. student Zhonghua Xi created several web-based tools for creating crease patterns and planning folding motions. (2014 Spring)

image
Predict Collisions
We found a way to compute the earliest collision time for a mobile robot (modeled as a point or polygon) moving among obstacles (polygon or articulated object) whose motion is unknown. (2013 Winter)


More in archived research hightlighs

All Projects



Research projects at MASC group are supported in part by NSF, DOT (FHWA), USGS, AFOSR, CIT and George Mason University

List of MASC Research Pages
Computer Science @ George Mason University