Connelly Barnes
510 Rice Hall
85 Engineer's Way, Box 400740
Charlottesville, VA 22904
connelly AT cs.virginia.edu
85 Engineer's Way, Box 400740
Charlottesville, VA 22904
connelly AT cs.virginia.edu
Research Interests and Bio
My group develops techniques that are on the boundary between computer graphics and computer vision, with a focus on graphics. We tend to incorporate in our research machine learning techniques such as non-parametric and recently, deep learning models. We focus on applications in computational photography, image and video synthesis, art, and hiding visual information. Many computer graphics algorithms are more useful if they are interactive, therefore, we also have a focus on efficiency and optimization, including some compiler technologies.
I am currently recruiting Ph.D. students so let me know if you are interested in applying to my lab (see admissions).
I received my Ph.D. from Princeton University in 2011. See my C.V. for more information.
I am currently recruiting Ph.D. students so let me know if you are interested in applying to my lab (see admissions).
I received my Ph.D. from Princeton University in 2011. See my C.V. for more information.
Classes
Computer Science 4501 - Introduction to Computer Vision (Spring 2017)
Computer Science 6501 - Deep Learning for Computer Graphics (Fall 2016)
See also a list of classes in previous years.
Computer Science 6501 - Deep Learning for Computer Graphics (Fall 2016)
See also a list of classes in previous years.
Impact and Press Coverage
The PatchMatch technology that I researched for my Ph.D. was incorporated into Adobe Photoshop CS5 as content-aware fill. This received 8 million views on YouTube [1] [2], and was featured in press such as Popular Science and PC Magazine. In later Photoshop versions, Content-aware Patch, Move, and Color Adaptation were built on this technology [3].
I collaborated on the Halide project, which has been used at Google. My collaboration on camouflage has been featured on Wired and Gizmodo, with 170,000 views on YouTube.
I collaborated on the Halide project, which has been used at Google. My collaboration on camouflage has been featured on Wired and Gizmodo, with 170,000 views on YouTube.
Graduate Students
- Yuting Yang (Ph.D.)
- Ning Yu (Ph.D.)
- Fuwen Tan (Ph.D.)
- Zachary Verham (Masters)
- Liming Lou (former visiting Ph.D.)
- ShanShan He (graduated Masters)
- Paul Nguyen (graduated Masters)
Research Publications (Google
Scholar)
VizGen: Accelerating Visual Computing Prototypes in Dynamic Languages
ACM SIGGRAPH Asia 2016
Halide: A Language and Compiler for Optimizing Parallelism, Locality, and Recomputation in Image Processing Pipelines
Communications of the ACM: Research Highlights (to appear)
Image Perforation: Automatically Accelerating Image Pipelines by Intelligently Skipping Samples
ACM Transactions on Graphics 2016 (to appear at ACM SIGGRAPH 2016)
Towards Automatic Band-Limited Procedural Shaders
Pacific Graphics 2015
PatchTable: Efficient Patch Queries for Large Datasets and Applications
ACM SIGGRAPH 2015
Synthesis of Complex Image Appearance from Limited Exemplars
ACM Transactions on Graphics 2015 (presented at ACM SIGGRAPH)
RealPigment: Paint Compositing by Example
NPAR 2014: Symposium on Non-Photorealistic Animation and Rendering
Stylized Keyframe Animation of Fluid Simulations
NPAR 2014: Symposium on Non-Photorealistic Animation and Rendering
Style Transfer for Headshot Portraits
ACM SIGGRAPH 2014
DecoBrush: Drawing Structured Decorative Patterns by Example
ACM SIGGRAPH 2014
Camouflaging an Object from Many Viewpoints
CVPR 2014 (oral presentation)
Patch-based High Dynamic Range Video
ACM SIGGRAPH Asia 2013
Halide: A Language and Compiler for Optimizing Parallelism, Locality and Recomputation in Image Processing Pipelines
ACM PLDI 2013
Image Melding: Combining Inconsistent Images using Patch-based Synthesis
ACM SIGGRAPH 2012
The PatchMatch Randomized Matching Algorithm for Image Manipulation
Communications of the ACM: Research Highlights, 2011
PatchMatch: A Fast Randomized Matching Algorithm with Application to Image and Video
Ph.D. Dissertation, Princeton University, 2011
PatchMatch: A Randomized Correspondence Algorithm for Structural Image Editing
ACM SIGGRAPH 2009
Video Puppetry: A Performative Interface for Cutout Animation
ACM SIGGRAPH Asia 2008
Education Publications
Enhancement of Student Learning in Experimental Design Using a Virtual Laboratory
IEEE Transactions on Education, 2008
Experiential Learning of Design of Experiments Using a Virtual CVD Reactor
American Society for Engineering Education Conference, 2006. Award for best paper in Chemical Engineering.
The Virtual CVD Learning Platform
Frontiers in Education Conference, 2006.
ThermoSolver: An Integrated Educational Thermodynamics Software Program
Undergraduate honors thesis, 2006. Oregon State University Library
Funding
Thanks for research funding from the following organizations: the National Science Foundation, Adobe, and the Chinese Scholarship Council.
Other Resources
My personal website includes other resources such as:
- Source code, including PatchMatch code and Python bindings for the Halide language.
- Artwork
- Poetry
- Music
- Writings
- Website mirror