Chang Yuan

ContactInformation

3100 SE 168th Ave #55 Phone: +1-360-450-5736Vancouver, WA 98683 E-mail: changyuan@gmail.comUSA WWW: http://iris.usc.edu/˜cyuan

ResearchInterests

Computer vision, image and video processing, computer graphics, human-computer interaction,virtual reality, and multimedia systems

Education University of Southern California, Los Angeles, California, USA

Ph.D. in Computer Science, August 2007∙ Advisor: Dr. Gerard G. Medioni

M.S. in Computer Science, May 2005

Tsinghua University, Beijing, China

B.Eng. in Computer Science and Technology, July 2001

ProfessionalExperience

Sharp Laboratories of America, Camas, Washington, USASenior Display Algorithm Researcher June 2007 - Present

Developed new algorithms and software for enhancing viewing experiences on LCD displays andautomatic production of LCD panels, including∙ Image synthesis algorithms for comfort retuning of stereoscopic content∙ Viewer tracking and distributed rendering algorithms for interactive viewing of ultra-high-resolution

visual content on tiled displays∙ Image enhancement and pattern matching algorithms and software for automatic inspection and

repair of defective LCD pixels (used in Sharp’s production lines)

University of Southern California, Los Angeles, California, USAGraduate Research Assistant August 2003 - May 2007

Proposed a novel approach to both 2D motion segmentation and 3D volumetric reconstruction ofmoving objects from a moving camera∙ Discovered a multi-frame 2D geometric constraint for detecting moving objects in presence of

motion parallax∙ Proposed a planar-motion constraint for computing 3D trajectories of moving objects without

scale ambiguity∙ Developed a volumetric reconstruction method to recover the 3D shape of both moving objects

and static background∙ Built a robust and efficient software system for 2D motion segmentation and 3D shape recon-

struction (patent pending)

Microsoft Research, Redmond, Washington, USASummer Intern May 2006 - August 2006

Designed a general framework for evaluating and improving stability of image interest points. Pro-posed a machine learning based approach to predict the stability of interest points from their localregions.

University of Southern California, Los Angeles, California, USAGraduate Teaching Assistant August 2003 - May 2004Responsibilities include teaching tutorial and discussion sessions, writing exam questions, and hold-ing office hours. Courses include:

1

∙ CSCI 351 (undergraduate level): Programming and Multimedia on the World Wide Web∙ CSCI 561 (graduate level): Foundations of Artificial Intelligence

Microsoft Research Asia, Beijing, ChinaVisiting Student (full-time intern) July 2002 - July 2003

Developed a novel method to extract spatial-temporal object trajectories from videos by mergingover-segmented 2D image regions based on their appearance and motion coherence.

Publications Referred Journal Publications

∙ Chang Yuan, Gerard Medioni, Jinman Kang, and Isaac Cohen, “Detecting Motion Regions inPresence of Strong Parallax from a Moving Camera with Multi-view Geometric Constraints,”IEEE Trans. on Pattern Analysis and Machine Intelligence (PAMI), vol. 29, no. 9, 2007

Referred Conference Publications

∙ Chang Yuan, “Using Large-size 2D Displays to Create 3D Hyper-Realistic See-Through Ex-periences,” to appear in the Proc. of 16th International Display Workshops, Miyazaki, Japan,12/2009

∙ Sachin Deshpande, Chang Yuan, Scott Daly, and Ibrahim Sezan, “A Large Ultra High Reso-lution Tiled Display System: Architecture, Technologies, Applications, and Tools,” to appear inthe Proc. of 16th International Display Workshops, Miyazaki, Japan, 12/2009

∙ Chang Yuan, “Creating Virtual 3D See-Through Experiences on Large-size 2D Displays,” inthe Proc. of IEEE Conf. on Virtual Reality, pp. 237-238, Lafayette, Louisiana, 03/2009

∙ Chang Yuan and Gerard Medioni, “Inferring 3D Volumetric Shape of Both Moving Objects andStatic Background Observed by a Moving Camera,” in the Proc. of IEEE Conf. on ComputerVision and Pattern Recognition (CVPR), pp. 1-8, 06/2007

∙ Chang Yuan and Gerard Medioni, “3D Reconstruction of Background and Objects Moving onGround Plane Viewed from a Moving Camera,” in the Proc. of IEEE Conf. on Computer Visionand Pattern Recognition (CVPR), vol. 2, pp. 2261-2268, 06/2006

∙ Jinman Kang, Isaac Cohen, Gerard Medioni and Chang Yuan, “Detection and Tracking ofMoving Objects from a Moving Platform in Presence of Strong Parallax,” in the Proc. of Intl.Conf. on Computer Vision (ICCV), vol. 1, pp. 10-17, 10/2005

∙ Chang Yuan, Yu-Fei Ma, and Hong-Jiang Zhang, “Extracting Video Object’s Motion Trajectoryby Velocity Voting,” in the Proc. of IEEE Pacific-Rim Conf. on Multimedia (PCM), vol. 3, pp.1561-1565, 12/2003

Thesis∙ Chang Yuan, “Motion Segmentation and Dense Reconstruction of Scenes Containing Moving

Objects Observed by a Moving Camera,” Doctoral dissertation, University of Southern California,05/2007

Talks /Presentations

∙ Poster presentation, IEEE Conf. on Virtual Reality, Lafayette, Louisiana, 03/2009∙ Poster presenation, IEEE Conf. on Computer Vision and Pattern Recognition, Minneapolis,

Minnesota, USA, 06/2007∙ Poster presenation, IEEE Conf. on Computer Vision and Pattern Recognition, New York City,

New York, USA, 06/2006∙ Oral presenation, Intl. Conf. on Computer Vision, Beijing, China, 10/2005

Patents ∙ Co-inventor with Louis Kerofsky, Sachin Deshpande, Xinyu Xu, and Scott Daly, “Super GreenTV System and Methods,” US Patent Pending, 2009

∙ Co-inventor with Robin Held, Hao Pan, Scott Daly, and Ibrahim Sezan, “Prevention of ViewerDiscomfort by Modifying the Presentation of Stereoscopic Image,” US Patent Pending, 2009

∙ Co-Inventor with Scott Daly, “Temporal Parallax Induced Display,” US Patent Pending, 2009

2

∙ Inventor, “System for Rendering Virtual See-through Scenes,” US Patent Pending, 2008∙ Co-inventor with Andrew Ferlitsch, “Methods and Systems for Imaging Device and Display

Interaction,” US Patent Pending, 2008∙ Co-inventor with Ratko Jagodic, and Sachin Deshpande, “A System for Distributed Widgets for

Large High Resolution Displays,” 2008∙ Co-inventor with Scott Daly and Dean Messing, “Method and System for Region based Upscal-

ing,” US Patent Pending, 2008∙ Co-inventor with Scott Daly, “Method and Apparatus for Building Compound-eye Seeing Dis-

plays,” 2007∙ Inventor, “Method and Apparatus for Rendering Virtual See-through Scenes on Single or Tiled

Displays,” US Patent Pending, 2007∙ Co-inventor with Gerard Medioni, Jinman Kang, and Isaac Cohen, “Detection and Tracking of

Moving Objects from a Moving Platform in Presence of Strong Parallax,” US Patent Pending,2007

ProfessionalServices

∙ Reviewer, IEEE Transaction on Pattern Analysis and Machine Intelligence (PAMI)∙ Reviewer, Journal of Zhejiang University Science∙ Program committee member, IEEE Workshop on Applications of Computer Vision (WACV),

2009∙ Reviewer, ACM International Conference on Multimedia, 2009∙ Program committee member, Workshop on Cloud-Mobile Convergence for Virtual Reality (CM-

CVR ’09), in conjunction with IEEE VR 2009∙ Reviewer, Workshop on 3D User Interface (3DUI ’09), in conjunction with IEEE VR 2009∙ Reviewer, 14th Intl. Conf. on Image Analysis and Processing (ICIAP), 2007

Honors andAwards

∙ Student Travel Award, IEEE Computer Society, 10/2005∙ Academic Achievement Award for International Students, University of Southern California,

05/2005∙ Outstanding Graduates of Tsinghua University, 07/2001∙ Tsinghua University Scholarship, 11/2000 and 11/1998

Memberships ∙ Member (since 1999), Institute of Electrical and Electronics Engineers (IEEE)∙ Member (since 2007), Association for Computing Machinery (ACM)∙ Member (since 2007), Society for Information Display (SID)

Skills ∙ Programming Languages: C/C++, Matlab, Python, Java, Lua∙ Programming Packages: OpenCV, IPP, OpenGL, OGRE 3D, VRPN, wxWidgets, FLTK∙ Operating Systems: Windows, Unix/Linux, Mac OS

References Available upon request

3