Richard Szeliski

Contact info:
Richard Szeliski
Affiliate Faculty
University of Washington

Publications (or try Google Scholar, MS Academic Search, or DBLP)

Prior to retiring, I was the founding Director of the Computational Photography group at Facebook.
I currently still have an Affiliate Professor appointment at the University of Washington.


Recent news:

The second edition of my computer vision textbook has finally been published. Please check it out and send me comments.

In December 2020, I retired from Facebook and am no longer actively working in computer science research.

In October 2017, I was awarded the PAMI Distinguished Researcher Award.

After twenty amazing years at Microsoft Research, where I led the Interactive Visual Media Group, I joined Facebook (in late October, 2015) as the founding Director of the Computational Photography group.

In February 2015, I was elected to the National Academy of Engineering.

Research areas:

My research areas include computer vision and computer graphics.

My particular interests are in using vision to automatically build 3-D models from images, computational photography, and image-based rendering. I have worked on both traditional 3-D volumetric and surface model reconstruction, and on high-resolution image mosaic construction. I am also interested in using computer vision for human-computer interaction and for analysing image databases. I have additional research interests in geometric modeling, motion estimation, multiresolution algorithms and representations, and optimization algorithms.

Benchmark studies:

Middlebury Stereo Vision Page (with Daniel Scharstein): an online evaluation of the best two-image stereo matching algorithms, along with test images, ground truth, and software.

MRF Energy Minimization Page (with Ramin Zabih, Daniel Scharstein, Olga Veksler, Vladimir Kolmogorov, Aseem Agarwala, Marshall Tappen, and Carsten Rother): a comparison of 2-D MRF energy minimization algorithms, along with test images, data files, and software.

Multi-View Stereo Evaluation (with Steve Seitz, Brian Curless, James Diebel, and Daniel Scharstein): a comparison of multi-view stereo 3D surface/volume reconstruction algorithms, along with calibrated test images.

Optical Flow Evaluation (with Simon Baker, Daniel Scharstein, JP Lewis, Stefan Roth, and Michael Black): a comparison of optic flow (motion estimation) algorithms, along with calibrated test sequences.


I was a founding editor of Foundations and TrendsĀ® in Computer Graphics and Vision. If you are interested in contributing a survey article, please contact one of the Editors-in-Chief.

Courses taught:

Computer Vision, UW CSE 576 ( Spring 2020, Spring 2008, Spring 2005)

Introduction to Computer Vision, Stanford CS223B, Winter 2003

Vision for Graphics, UW CSE590SS, Winter 2001

NIPS 2004 Tutorial on Acquiring Detailed 3D Models From Images and Video

ICCV 2003 Short Course on Dense multiview stereo

ICCV 2003 Short Course on Image-based Rendering


Photosynth was a publicly available service based on our original Photo Tourism paper

The Image Compositing Editor (ICE) is an advanced image stitching tool.

Places I have been affiliated with:

University of Washington Graphics and Imaging Laboratory

The Computational Photography group at Facebook

The Interactive Visual Media Group at Microsoft Research

Digital Equipment Corporation's Cambridge Research Laboratory

SRI International's AI Center (Perception Program)

Carnegie Mellon Vision and Autonomous Systems Center     (Computer Vision Home Page)


Last updated 1/19/2022