Yoni Wexler's Home Page

I am a researcher in the Virtual Earth Group at Microsoft. My research interests lie in the field of Computer Vision, Geometry in Vision (Motion Analysis, Shape Description) for perceptual tasks such as Action Recognition, View Synthesis and Augmented Reality. My work has been concentrated particularly on the recovery and representation of scenes from images. My research focuses on finding computational methods that enable accurate and efficient use of pictorial data, both in images and video sequences.

Space-Time Video Completion

Space-Time Scene Manifolds

Alpha Matting

Environment Matting

Qwarping

Dynamic Morphing

Publications

Factoring repeated content within and among images.
H. Wang, Y. Wexler, E. Ofek, H. Hoppe.
ACM SIGGRAPH 2008 , pdf

Hierarchical photo organization using geo-relevance
B. Epshtein, E. Ofek, Y. Wexler and P. Zhang
ACM GIS 2007, Seattle, Washington. abstract, pdf

Space-Time Scene Manifolds
Y. Wexler and D. Simakov
International Conference on Computer Vision (ICCV) 2005, abstract , pdf

Space-Time Video Completion
Y. Wexler, E. Shechtman, and M. Irani
In IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), March 2007. pdf.

Previously appeared in Computer Vision and Pattern Recognition (CVPR) 2004 and received Best Poster Award.  abstract, pdf

Image-Based Rendering Using Image-Based Priors
A.W. Fitzgibbon Y. Wexler, and A. Zisserman.
International Conference of Computer Vision (ICCV) 2003. Received the David Marr Prize.
Abstract, pdf, Data
Also appeared in the International Journal of Computer Vision, 63(2):141–151, July 2005. Marr Prize special issue.

Learning Epipolar Geometry from Image Sequences
Y. Wexler, A.W. Fitzgibbon and A. Zisserman
IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), June 2003
abstract, pdf

Image-based Environment Matting
Y. Wexler, A. Fitzgibbon and A. Zisserman.
Proc. 13th Eurographics Workshop on Rendering, Italy, 2002. pdf
Shorter version appeared as SIGGRAPH 02' sketch, San Antonio. pdf

Bayesian Estimation of Layers from Multiple Images.
Y. Wexler, A. Fitzgibbon and A. Zisserman.
Proc. 7th European Conference on Computer Vision (ECCV), Copenhagen, Denmark, 2002.
ps.gz, pdf.

View Synthesis using Convex and Visual Hulls
Y. Wexler, R. Chellappa.
In Proc. British Machine Vision Conference 2001', Manchester.
pdf

Join Tensors: On 3D-to-3D Alignment of Dynamic Sets
L. Wolf, A. Shashua and Y. Wexler
Proc. of Int. Conf. on Pattern Recog. (ICPR), Sep. 2000, Barcelona. pdf.
Also, CfAR report 4097, January 2000: ps.gz pdf

On the Synthesis of Dynamic Scenes from Reference Views.
Y. Wexler and A. Shashua
IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), June 2000, South Carolina
CfAR report 4096, January 2000. ps.gz pdf
Sample output automatically builds this movie from these 3 images.

Q-warping: Direct Computation of Quadratic Reference Surfaces.
Y. Wexler and A. Shashua.
IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), June 1999. pdf
Also, PAMI journal version appeared in August 2001 (Vol. 23, No. 8, pp. 920-925) pdf

The Marr Prize Best Paper Award at the 9th International Conference of Computer Vision (ICCV) 2003

Best Poster Award at the 22nd IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2004)