Yasutaka Furukawa

http://www.cs.washington.edu/homes/furukawa/

 

Athabasca Glacier, Banff, Canada

Yasutaka Furukawa Post-doctoral Research Associate Department of Computer Science and Engineering University of Washington Box 352350, Seattle, Washington 98195-2350, USA Office: Allen Center 276 Email: click here Phone: +1-206-543-3069 Fax: +1-206-543-2969

I work for Google now, while I am still involved in some academic projects. I hope to provide exciting products to people in the world and prove that computer vision is exciting and useful.

---

What's new

ECCV 2012 paper accepted (oral).

Google Maps - Photo Tours launched! (2012/04)

Google MapsGL launched! (2011/10)

INTEL CEO talked about our itnernet-scale MVS work

PMVS is used for real film production by Industrial Light & Magic.

---

Who Am I ?

I am a post-doctoral research associate (or postdoc) in the Graphics and Imaging Laboratory at University of Washington. I work with Prof. Seitz and Prof. Curless at University of Washington, and Rick Szeliski at Microsoft Research. I obtained a B.S. degree under the supervision of Dr. Yoshihisa Shinagawa at University of Tokyo in March 2001. I completed my Ph.D. under the supervision of Prof. Ponce at Computer Science Department of Univeristy of Illinois at Urbana-Champaign in May 2008. My primary research interests are in Computer Vision and Computer Graphics. The following is a list of projects I've worked on.

CV (and links to all the publications)

---

Research

	Reconstructing the World's Museums Jianxiong Xiao and Yasutaka Furukawa To appear at ECCV 2012 (oral).
	Google Maps - Photo Tours, April, 2012 Designed and implemented a pipeline to compute scene geometries for rendering and "3D Ken Burns" motions. Examples (a few clicks away).
	MapsGL (Next generation Google Map), October, 2011 Designed and implemented a city-scale dense 3D reconstruction pipeline from aerial images, running in an unprecedented scale for hundreds of cities all over the world.
	Building Rome in a Day Sameer Agarwal, Yasutaka Furukawa, Noah Snavely, Brian Curless, Steven M. Seitz and Richard Szeliski Communications of the ACM, Vol. 54, No. 14, Pages 105-112, October 2011. Building Rome Sameer Agarwal, Yasutaka Furukawa, Noah Snavely, Brian Curless, Steven M. Seitz and Richard Szeliski IEEE Computer Society, June 2010.
	Towards Internet-scale Multi-view Stereo (supplementary material) Yasutaka Furukawa, Brian Curless, Steven M. Seitz and Richard Szeliski CVPR 2010. CMVS software available! Project website coming soon.
	Reconstructing Building Interiors from Images Yasutaka Furukawa, Brian Curless, Steven M. Seitz and Richard Szeliski ICCV 2009 (oral). Project website
	Dense 3D Motion Capture for Human Faces Yasutaka Furukawa and Jean Ponce CVPR 2009. Movie
	Manhattan-world Stereo Yasutaka Furukawa, Brian Curless, Steven M. Seitz and Richard Szeliski CVPR 2009. Movie
	Dense 3D Motion Capture from Synchronized Video Streams Yasutaka Furukawa and Jean Ponce CVPR 2008. (oral) Dense Patch Models for Motion Capture from Synchronized Video Streams Yasutaka Furukawa and Jean Ponce -- Willow Tech. Report 02-07 Movie (Also see our gallery for more movies)
	Accurate Camera Calibration from Multi-View Stereo and Bundle Adjustment Yasutaka Furukawa and Jean Ponce IJCV, 2009. Accurate Camera Calibration from Multi-View Stereo and Bundle Adjustment Yasutaka Furukawa and Jean Ponce CVPR 2008. Movie
	Accurate, Dense, and Robust Multi-View Stereopsis Yasutaka Furukawa and Jean Ponce PAMI, August 2010. Accurate, Dense, and Robust Multi-View Stereopsis Yasutaka Furukawa and Jean Ponce CVPR 2007. High-Fidelity Image Based Modeling Yasutaka Furukawa and Jean Ponce -- Technical Report 2006-02 Multi-View Stereo Software for download. See our gallery and 3D Photography Datasets. See the multi-view stereo evaluation project by Steve Seitz et al. for quantitative evaluations of our results. As of 9/25/2007, our algorithm achieves the best result in terms of both accuracy and completeness for 4 out of 6 datasets.
	Carved Visual Hulls for Image-Based Modeling Yasutaka Furukawa and Jean Ponce IJCV, March 2008 Special Issue: Best of the European Conference on Computer Vision 2006 Carved Visual Hulls for Image-Based Modeling Yasutaka Furukawa and Jean Ponce ECCV 2006. (oral) Carved Visual Hulls for High-Accuracy Image-Based Modeling Yasutaka Furukawa and Jean Ponce -- SIGGRAPH 2005 Technical Sketch See our 3D Photography Datasets and visual hull data sets.
	Projective Visual Hulls IJCV, August 2007 Svetlana Lazebnik, Yasutaka Furukawa and Jean Ponce See our visual hull data sets.
	Robust Structure and Motion from Outlines of Smooth Curved Surfaces Yasutaka Furukawa, Amit Sethi, Jean Ponce, and David Kriegman. PAMI, February 2006 (Vol. 28, No. 2) pp.302-315 Structure and Motion from Images of Smooth Textureless Objects Yasutaka Furukawa, Amit Sethi, Jean Ponce, and David Kriegman. ECCV 2004.
	Accurate and Robust Line Segment Extraction by Analyzing Distribution around Peaks in Hough Space Yasutaka Furukawa and Yoshihisa Shinagawa Computer Vision and Image Understanding Volume 92, Issue 1, Pages 1-25 (October 2003)

---

Resources

• Changchang Wu distributes an automated end-to-end 3D reconstruction system from images, which integrates well with our CMVS and PMVS.
• I recently found a very interesting website, which has nice video tutorials on an end-to-end 3D reconstruction system (including PMVS2) on Windows (32/64 bits) machines. Here is the link.
• CVPR 2010 Tutorial on 3D shape Reconstruction from Photographs: a Multi-View Stereo Approach.
• Clustering software for Multi-view Stereo (CVPR 2010) is now available!
• Windows binary (32 and 64 bits versions) and modified source codes of PMVS2 is now available at Pierre Moulon's website!
• 3D viewer with the gallery dataset for download (ICCV paper)
• 2nd version of PMVS is now available with the source codes
• Gallery
• Patch-based bundle adjustment software(PBA) for download
• 3D Photography Datasets
• Visual hull data sets