Electrical Engineering

Yiming Li

Graduate Student Researcher, Center for Research in Intelligent Systems

Electrical Engineering Department, University of California, 216 Engineering Bldg. II, Riverside, CA 92521

Phone: 951-***-**** E-mail: *****@**.***.*** URL: http://www.ee.ucr.edu/~yimli

OBJECTIVE A highly motivated and qualified Ph. D candidate seeking for a full time research

or engineering position in computer vision, image processing, pattern recognition

and machine learning.

5 years research experience on computer vision, image processing, pattern recognition

SUMMARY

and machine learning.

10+ publications (conferences, journals and book chapters) in related area.

Solid knowledge in feature detection/extraction/description, image-based object

detection/tracking/recognition, camera calibration/multi-view geometry, IP camera

networks.

Knowledge in augmented reality and driving assistance systems (vehicle detection,

lane position tracking).

Proficient in object-oriented programming in C++ with STL, OpenCV, POSIX and

Boost (socket and multithreading), Matlab, Python, and HTML. Experience on source

control tools (Mercurial). Experience on Windows platforms.

Self-motivated, passionate, quick learner, strong demonstration and communication

skills, excellent organizational skills with attention to details, multi-task and

prioritizing tasks, good team player and pleasant to work with.

Ph.D. Candidate, Electrical Engineering, GPA 4.0/4.0

EDUCATION

Sept 07 - Present University of California, Riverside

M.S., Electrical and Computer Engineering,

Aug 06 Aug 07

University of Florida

B.S., Electrical Engineering,

Sept 02 - June 06

South China University of Technology

PROFESSIONAL Graduate Student Researcher, University of California, Riverside

Advisor: Dr. Bir Bhanu, Center for Research in Intelligent Systems

EXPERIENCE

Sept 07 - Present Camera selection and active control for VideoWeb, a wireless IP camera network

Introduced a series of classic models from economics to camera hand -off, selection and

active control problems. Solved the problems in a more flexible and computationally

effective way. Made large scale video networks more intelligent. Implemented a real-time

fully automatic video surveillance platform in C++ and OpenCV. Utilized multithreading

and TCP/IP socket programming in Boost.

Nov 10 Dec 11 Coupled Camera Selection and Object Tracking

Proposed a closed-loop framework for doing object tracking and camera selection

simultaneously. Developed a score-level fusion of multiple trackers which can be

benefited from camera selection results. Camera selection criteria, such as object position,

size, view, and spatial/temporal tracker smoothness, are considered together with the

performance of multiple trackers. The process of camera selection is integrated with

fusion of multiple trackers.

Built up a real-time framework. Simulated distributed systems by multi-threads.

Applied a series of state-of-the-art trackers, including online boosting tracker, semi -

supervised online boosting tracker, multiple instance learning tracker, P-N learning

tracker, CamShift tracker, and particle filter tracker.

Calibrated the experimental environment consisted of Axis 215 PTZ cameras.

Calculated homographies between camera pairs. Used epipolar geometry to help to locate

the object position and build up correspondences among cameras.

Sept 09 Dec 10 Camera Active Control by an Auction-based Model

Developed the idea of applying auction-based approaches in an economic market to

camera assignment problems. Proposed the concept of bidding vectors to consider the

potentially available cameras to track an object. Integrated camera active/PTZ controls

with camera hand-off and achieved it in a calibrated environment.

Oct 08 - Sept 09 Camera Selection in a Video Network by Weakly Acyclic Games

Modified the potential game model to the weakly acyclic game model to get

rid of the requirement of utility alignments. Made the approach more efficient and

the utility design more flexible. Used Payoff-based learning algorithm to get the

optimal as well as stable solution to the camera assignment problem.

Sept 07 - Oct 08 Camera Selection in Large-Scale Video Network by Potential Games

Created the novel idea of applying game theory to camera selection problems.

Matched the problem to the vehicle-target model as a potential game. Applied bargaining

mechanism to get the stable Nash Equilibrium as the final solution and proved for the

convergence. Designed several task-oriented criteria as the benchmarks to evaluate the

performance of the proposed approach.

Used the Viola & Jones approach and the HOG feature in OpenCV to do object

detection. Applied CamShift and particle filter trackers for object tracking. Established a

framework for video surveillance using Axis IP cameras. Deployed communication with

IP cameras by using socket programming provided by the Boost library.

Jun 09 Sept 09 Research Internship - ObjectVideo, Reston, Virginia

Cross camera vehicle re-identification

Used the PHOG feature and the SVM classifier to do vehicle detection.

Applied Procrustes analysis and principal component analysis (PCA) for 3D

vehicle model alignment.

Compared several techniques (Appearance Invariant Model, Ensemble of

Localized Features and etc.) for similarity computations to re-identify objects. Used the

real-world vehicle dataset collected by ObjectVideo and the VIPer pedestrian dataset

collected by UCSC.

Marketing Engineer Internship Santachi Video Technology, Shenzhen, China

Apr 06 Jun 06

Key presenter at 3 stops of one of the largest commercial events of Stanchi Video

Technology for new product releasing and sales training.

Product representative at ISC (International Security Conference & Exposition) East,

New York City, NY, 2009.

Product representative at ISC West, the world s largest exhibition in the video

surveillance industry, Las Vegas, NV, 2008.

RELEVANT Advanced Computer Vision, Advance Digital Image Processing, Pattern Recognition,

COURSES Current Topics on Computer Vision and Pattern Recognition, Advanced Data Structure,

Aug 06 Jun 10 Intelligent Systems, Stochastic Process, Statistical Methods in Research, Linear Systems.

Recipient of Best Paper Award, 2nd IEEE International Conference on Distributed

HONORS /

AWARDS Smart Cameras, Palo Alto, CA, Sept 7-11, 2008.

Recipient of Graduate Division Fellowship Award, University of California at

Riverside, 2007.

Recipient of Achievement Award of University of Florida, 2006.

Recipient of outstanding student scholarships, 2002-2006.

PROFESSIONAL Poster presentation, Workshop on Distributed Video Sensor Networks, UC Riverside

ACTIVITIES Talk on the colloquium of the Electrical Engineering Department, UC Riverside

Reviewer for Elsevier Journal of Pattern Recognition

Reviewer for Springer Journal of Machine Vision and Applications

Reviewer for IEEE Sensors Journal

Reviewer for Recent Patents on Electrical Engineering

Yiming Li and Bir Bhanu, Utility-Based Camera Assignment in a Video Network: A

PUBLICATIONS

Game Theoretic Framework, IEEE Sensors Journal, May 2010.

Yiming Li, Bir Bhanu, Fusion of Multiple Trackers in Video Networks, IEEE/ACM

International Conference on Distributed Smart Cameras, Ghent, Belgium, August, 2011.

Yiming Li, Bir Bhanu and Wei Lin, Auction Protocol for Camera Active Control,

IEEE International Conference on Image Processing, Hong Kong, China, September,

2010.

Yiming Li, Bir Bhanu and Vincent Nguyen, Evaluating Camera Handoff Techniques

on Different Trackers, IEEE International Conference on Pattern Recognition, Istanbul,

Turkey, August, 2010.

Yiming Li and Bir Bhanu, Task-Oriented Camera Assignment in a Video Network,

IEEE International Conference on Image Processing, Cairo, Egypt, November, 2009.

Yiming Li and Bir Bhanu, A Comparison of Techniques for Camera Selection and

Handoff in a Video Network, IEEE/ACM International Conference on Distributed Smart

Cameras, Como, Italy, September, 2009.

Yiming Li and Bir Bhanu, Utility-Based Dynamic Camera Assignment and Hand-off in a

Video Network, IEEE/ACM International Conference on Distributed Smart Cameras,

Stanford University, CA, USA, September, 2008. Best Paper Awarded.

Bir Bhanu and YimingLi, A Comparison of Techniques for Camera Selection and

Handoff in a Video Network, BirBhanu, Chinya V. Ravishankar, amit K. Roy-Chowdhury,

DemetriTerzopoulos and Hamid aghajan, Distributed Video Sensor Networks, Springer, May

2010.

Bir Bhanu and Yiming Li, A Game Theoretic Framework for Dynamic Camera

Assignment and Hand-off in a Video Network .Yunqian Ma and Gang Qian, Intelligent Video

Surveillance: Systems and Technology, Auerbach Publications, December, 2009.

Bir Bhanu and Yiming Li, Camera Assignment and Hand-off . Yan Zhang and Mohsen

Guizani, Game Theory for Wireless Communications and Networking, Auerbach Publications,

June, 2011.

Contact this candidate