IEEE Workshop on
Statistical Analysis in Computer Vision
In conjunction with CVPR 2003

---

Goals and Scope

Statistical analysis has always played an important role across a wide range of computer vision topics. The complexity and ill-posed nature of many computer vision problems, however, often makes statistical analysis using traditional approaches difficult. The goal of this workshop is to stimulate research in all aspects of statistical analysis in computer vision.

The workshop is aimed at bringing together researchers from computer vision, mathematics, statistics, information theory, and visual psychology to address key issues in statistical analysis in vision systems. Original papers are solicited on topics including, but not limited to:

• the use of advanced statistical tools in modeling, design,and evaluation of algorithms, systems, and sensors
• the propagation of errors models through vision modules
• the analysis and justification of statistical assumptions used in vision algorithms
• novel statistical methods applied to vision problems
• unification, generalizations, and extentions of existing statistical models
  

Research regarding the theoretical aspects of the above topics is especially encouraged. The expectation is for more detailed discussion of the "statistical" aspects than would occur in a normal CVPR paper.

Tentative Program

The following program is subject to minor modifications. Papers will be presented by the authors in 40 minute intervals, with a five minute cushion between talks to allow for Q&A, and setup/breakdown. An asterix indicates a short (20 minute) talk.

	Morning Session
8:50 to 9:00	Introductory Remarks
9:00 to 9:40	Statistical Error Propagation in 3D Modeling From Monocular Video Amit Roy Chowdhury, Rama Chellappa
9:45 to 10:25	Parametric Subpixel Matchpoint Recovery with Uncertainty Estimation: A Statistical Approach Matt Steele, Christopher Jaynes Break
10:45 to 11:25	Statistically Robust Approach to Lens Distortion Calibration with Model Selection Moumen Taha El-Melegy, Aly A. Farag
11:30 to 12:10	Statistical Models for Skin Detection Bruno Jedynak, Huicheng Zheng, Mohamed Daoudi Afternoon Session
1:00 to 1:40	Hierarchical Learning of Optimal Linear Representations Qiang Zhang, Xiuwen Liu, Anuj Srivastava
1:45 to 2:25	Higher-order Wavelet Statistics and their Application to Digital Forensics Hany Farid, Siwei Lyu Break
2:45 to 3:20	A Study on Bayesian Feature Fusion for Image Classification X. Shi, R. Manduchi
3:30 to 4:10	A Statistical Assessment of Subject Factors in the PCA Recognition of Human Faces Geof Givens, J. Ross Beveridge, Bruce A. Draper, David Bolme
4:15 to 4:55	Robust LDA Classification by Subsampling* Sanja Fidler, Ales Leonardis
4:55 to 5:05	Closing Remarks

Review and Publication Information

The organizing committee recieved 28 submissions for this workshop, out of which, eight were selected for 40 minute talks, and one was accepted for a 20 minute talk. Paper reviews were performed double-blind. Accepted papers will be published by the IEEE in CD-ROM format and indexed in the IEEExplore website.

If you have a question or comment about the workshop that is not already addressed by this web page, please e-mail Ross J. Micheals at rossm@nist.gov or Terry Boult at tboult@cse.lehigh.edu.