home
research
recent publications
courses
research - students
photos
curriculum vitae


Roger S. Gaborski

Rochester Institute of Technology

102 Lomb Memorial Drive Rochester, New York  14623-5608


Office: Golisano College of Computing and Information Sciences, Bldg 70, Room 3647
Phone: (585)-475-4931
Fax: (585)-475-7100
E-mail: rsg@cs.rit.edu
Homepage: www.cs.rit.edu/~rsg

 

Current Position

Professor, Department of Computer Science
Extended Faculty Member, Center for Imaging Science
 

Education

Ph.D.

Electrical Engineering

University of Maryland

M.S.

Electrical Engineering

SUNY at Buffalo

B.S.

Electrical Engineering

SUNY at Buffalo

Research Interests


Teaching Experience

4005-759 Biologically Inspired Intelligent Systems
4005-755 Neural Networks and Machine Learning
4005-757 Computer Vision
4005-759 Advanced Computer Vision
4005-750 Artificial Intelligence
4003-590 Special Topics in AI - Distributed Agents
4005-891 Graduate Project Seminar
1051-784 Spatial Pattern Recognition
4003-231 Computer Science 1
4003-232 Computer Science 2
4003-351 Introduction to Digital Design

Current Research Activities

Automatic Object Tracking and Interpretation in Video Sequences (AOTIVS)

The AOTIVS project has two major objectives. The first objective of this project is to detect, track and classify both salient and non-salient objects in video sequences. The second objective is to model and interpret the object’s behavioral activities. The objects may include individuals, small groups of individuals, crowds and vehicles. Modeled behaviors will be used as a video indexing method.  During the first phase of the project a stationary camera will be used to collect the video sequences. During the second phase networks of collaborating cameras and non-stationary cameras will be used.

VENUS – A System for novelty Detection in Video Streams with Learning

The VENUS (Video Exploitation and Novelty Understanding in Scenes) framework is hierarchical in structure and is comprised of both low level perceptual and higher level learning and habituation components. As a result of our everyday visual interactions and our analytical abilities, we acquire a general understanding of the environment and are able to distinguish between consistent and inconsistent events based on prior knowledge. Any event that does not fit our definition of normalcy tends to be recognized and classified as a novel event. If such an event occurs often, the event is ignored as a consequence of habituation and is no longer considered novel. In this activity, we extend this theory of habituation to include learning of what is and what is not a novel event for a computer vision system. Astronomical observations, surveillance systems, ultrasound and cardiac medical imaging are examples of systems that generate streams of video data. It is very tedious for a human analyst to observe and interpret all this data and find “novel events”. In order to manage and eventually utilize the information extracted from video, it is necessary to develop a framework that can extract meaningful semantic information such that it reflects the description understood by a human analyst. As a simple example, consider a scene in which cars travel along a road in a right to left direction. The system will automatically learn the environment, that is, the movement of cars in a given region of the visual field. If a car now enters the scene, but this time travels from left to right, the directional movement of the car would be considered inconsistent. It is also possible for an object to be inconsistent in attributes such as, color, speed, location, etc.

The goal of this project is to develop the foundations for a computational learning framework for novelty detection, habituation and understanding based on biologically inspired principles. Such a framework will enable us to gain a better understanding of the issues and challenges associated with extracting a semantic understanding from unstructured video data.

Goal Directed Visual Search Based on Color Cues:
Co-operative Effects of Top-Down & Bottom-Up Visual Attention

The goal of this project is to develop a model of certain functions of the human visual system. Focus of attention is a critical mechanism we adopt while observing our surroundings. We, as humans, do not perceive every aspect of the surroundings, but instead focus on the interesting aspects and ignore the non-interesting ones. Thus, attention acts as a gating mechanism to the higher level processing in the visual cortex, processing only attended locations of the scene. This prioritizing of objects leads us to the concept of saliency based on low-level features of objects. These low level properties of objects are extracted by the bottom up focus of attention initially described by Kock & Ullman (1985), where a saliency map reflects the relative conspicuity of objects from their surroundings. In this project we develop multiresolution (multiscale) intensity difference, color opponent and orientation features. The intensity difference and color opponent features are implemented using receptive features. The orientation filters are modeled using oriented Gabor functions. A top-down cognitive process modulates the earlier bottom up processing. For instance, given a visual search task, the knowledge about the target’s features biases the perception of the scene. High-level knowledge about the object’s features may include shape, color or motion information. In this project we first investigate a saliency map based on bottom up features. In the second phase of the project the low level saliency map of natural images are modified by top down color cues. A top-down system will be implemented using a neural network modeling the working memory area of the brain and the bottom up system is implemented using conspicuity maps.

Steganography and Steganalysis

Steganography is the processing of ‘hiding’ information in images. The goal of this project is to investigate published steganographic techniques and evaluate their effectiveness over a wide range of image conditions. A second goal is the development of  steganalysis algorithms to detect if data is hidden in an image or not.


Professional Organization Membership


Recent Professional Activities


Consulting Activities

Recent consulting arrangements in the areas of computer vision, machine learning, image understanding and neural networks:


Publications, Patents, Academic Grants and Awards 



Professional Employment History

Graduate Coordinator, Department of Computer Science

     - Responsible for the operation of the graduate program in Computer Science

Adjunct Faculty, Department of Computer Science and Center for Imaging Science (RIT)
    - Taught neural network and pattern recognition courses

Eastman Kodak Company, Senior Research Associate

Worked directly with the Director of New Business Opportunities and reported to the Vice President, Health Imaging Division on technical developments.

-         Managed a group of researchers developing medical imaging products. Worked with customers to obtain input for product improvements.
-         Developed image processing algorithms for digital medical imaging products. The algorithms included image segmentation, image understanding, enhancement,         tone scaling and compression.
-         Develop an image understanding and visualization system for medical images. The techniques employed included shape understanding, data fusion techniques and     a probabilistic reasoning unit that incorporated anatomical knowledge.
-         Developed algorithms for computed aided mammography.
-         Developed cooperative research programs with University of Rochester and SUNY at Buffalo.

 

Recruited and lead a team of researchers from several disciplines to develop document understanding and optical character recognition algorithms.

-         Responsible for providing direction in document understanding issues, including automatic form recognition and optical character recognition (machine print, hand print and cursive).
-         Developed cooperative research programs with research groups at SUNY at Buffalo and MCC (Microelectronic and Computer Consortium, Austin, TX )
 
Member of a select team whose purpose was to formulate the research directions in the area of image understanding for the Eastman Kodak Company


National Security Agency

   - Developed algorithms for speech understanding
   - Developed special purpose signal processing algorithms and hardware

Calspan
    - Developed low power instrumentation to detect oil pollution in waterways. This work involved utilizing optical sensors and analog/digital circuit design

 

Presentations and Publications

 
Teredesai, A., Kanodia, J., Gaborski, R. and Ahmad, M., “COMMA: Combing Multimedia Multi-relational Associations,” (The 7th ACM SIGKDD Workshop on Multimedia Data Mining, Seattle, WA, USA, August 2004).

Gaborski, R., Vaingankar, A, Chaoji, V. and Teredesai, A., "VENUS: A System for Novelty Detection in Video Streams with Learning," (FLAIRS, May, 2004).
 
Gaborski, R., Vaingankar, A. and Tentler, A., "Detection of Inconsistent Regions in Video Streams," (SPIE January, 2004).
 
Vaingankar, A, Chaoji, V., Gaborski, R. and Teredesai, A., “Cognitively Motivated Habituation for novelty Detection in Video”, (NIPS 2003 workshop on Open Challenges in Cognitive Vision, December, 2003.

Gaborski, R., Vaingankar, A. and Cansova, R., "Goal Directed Visual Search Based on Color Cues: Co-operative Effects of Top-Down & Bottom-Up Visual Attention," (ANNIE, 2003, November, 2003)

Brazeau, J., Humphrey, J. and Gaborski, R., "A Biologically-based Model of a Visual Saliency Model," (IEEE Western NY Workshop, September, 2003)

Tentler, A., Vaingankar, A., Gaborski, R. and Teredesai, A., "Event Detection in Video Sequences of Natural Scenes," (IEEE Western NY Workshop, September, 2003)

Keller, J., Gaborski, R., Vaingankar, A. and Tentler, A. and Tymann, P., "Parallel Simulation of a Visual Saliency Model," ((IEEE Western NY Workshop, September, 2003)

Hollinger, D., Anderson, P., Gaborski, R. and Teredesai, A., “Feature Extraction and Classification in Machine Printed Korean OCR,” (IEEE Western NY Workshop, September, 2003)

Pakin, S.K., Gaborski, R., Barski, L., Foos, D. and Parker, K. J., "A Clustering Approach to Bone and Soft Tissue Segmentation of Digital Radiographic Images of Extremities,” Journal of Electronic Imaging, January, 2003.

Luo, H., Acharya, R. and Gaborski, R., "Knowledge-based Image Understanding and Classification System for Medical Image Databases",  SPIE International Symposium on Medical Imaging 2002.

Luo, H., Gaborski, R. and Acharya, R., "Knowledge Representation for Image Content Analysis in Medical Image Databases", SPIE International Symposium on Medical Imaging 2001.

Pakin, S.K., Gaborski, R., Barski, L., Foos, D. and Parker, K. J., "Segmentation Of Bone And Soft Tissue Regions In Digital Radiographic Images Of Extremities,” SPIE International Symposium on Medical Imaging 2001.

Pakin, S.K., Gaborski, R., Barski, L., Foos, D. and Parker, K. J., "An Algorithm for Segmentation of Bone and Soft Tissue in X-ray Extremity Images", 2000 Western New York Image Processing Conference Workshop, Rochester, New York, 2000.

Luo,H., Gaborski, R. and Acharya, R.,  "Automatic Segmentation of Lung Regions in Chest Radiographs: A Model Guided Approach", IEEE International Conference on Image Processing (ICIP2000), Vancouver, Canada,  2000.

Luo, H., Gaborski, R. and Acharya, R., "Robust Snake Model", Computer Vision and Pattern Recognition 2000, CVPR2000, Hilton Head Island, SC., 2000.

Luo, H., Acharya, R.  and Gaborski,R. , " A Knowledge-based Method for Automatic Segmentation of Lung Regions in Digital Chest Radiographs", 1999 Western New York Image Processing Conference Workshop, Rochester, New York, 1999.

Sun, Y., Gaborski, R. and Acharya, R., "A Practical Approach for Locating Bone Structures in Radiographic Pelvis Images",  1999 Western New York Image Processing Conference Workshop, Rochester, New York, 1999.

Pawlicki, T and Gaborski, R., "Identifying Phangeal Regions in Digital X-ray Images", International congress on Computational Intelligence Methods and applications, Rochester, New York, 1999.

Gaborski, R. and Oswalt, E., "Locating People in Images", International congress on Computational Intelligence Methods and applications,  Rochester, New York, 1999.

Luo, H., Acharya, R.  and Gaborski,R. , " Fully Automatic Detection of Spine in Chest Radiographs using Fuzzy Logic Approach, "  Soft computing in Biomedicine, Rochester, New York, 1999.

Luo, H., Acharya, R., Gaborski, R., " A New Fully Automatic Approach to Detect the Spine from X-ray Image", IEEE Western New York Image Processing Workshop, Rochester, New York, 1998.

Cook, L.T., Cox, G.G., Insana, M.F., McFadden, M.A., Hall, T.J., Gaborski, R.S., Lure, F.Y.M., "Comparison of a Cathode-Ray-Tube and Film for Display of Computed Radiographic Images,"  Medical Physics , vol. 25, pp. 1132-1138, 1998.

Gaborski, R.S., "A Knowledge Based Approach for Landmark Detection in Medical Images," 1997 Western New York Image Processing Workshop, Rochester, New York, 1997.

Cook, L.T., Cox, G.G., Insana, M.F., McFadden, M.A., Hall, T.J., Gaborski, R.S., Lure, F.Y.M. "Contrast-detail analysis of the effect of image compression on computed tomographic images," , Proceedings of the SPIE: Image Perception , Vol. 2712, pp. 128-137, 1996.

Lure, F., Pawlicki, T., and Gaborski, R., "Improvement of Detection Accuracy in Digital Mammography with a Pruned Neural Net Optimized from Heuristic Decision Rules," ICNN 96, Washington, DC, 1996.

Lure, F., Gaborski, R., and Pawlicki, F., "Application of Neural Network-based Multistage System for the Detection of Microcalcification in Mammogram Images," SPIE Medical Imaging 1996, Newport Beach, CA. 1996.

Lure, F., Jones, P., and Gaborski, R., "Multiresolution  Unsharp Masking Technique for Mammogram Image Enhancement," SPIE Medical Imaging 1996, Newport Beach, CA. 1996.

Wen, C., Lure, F., and Gaborski,R., "Evaluation of the Diagnostic Quality of Chest Images Compressed with JPEG and Wavelet Techniques," SPIE Medical Imaging 1996, Newport Beach, CA. 1996

Cook, L., Cox, G., Insana, M., McFadden, M., Hall, T., Gaborski, R., and Lure, F., ?Contrast-detail Analysis of the Effect of Image Compression on Electronically Displayed Computed Radiographic and Computer Tomographic Images,? SPIE Medical Imaging 1996, Newport Beach, CA. 1996

Wen, C. and Gaborski, R., “Image Quality Evaluation of Chest Radiographs Compressed with JPEG and Wavelet Algorithms,” Sixth Far West Perception Conference, Philadelphia, PA. 1995.

Anderson, P., Gaborski, R., Ge, M., Raghavendra, S., and Lung, M-L., “Using Quasirandom Numbers in NN”, Proceedings of the International ICSC Symposium on FUZZY LOGIC, Swiss Federal Institute of Technology (ETH), Zurich, Switzerland, 1995.

Gaborski, R.S. and Lure, F.Y.M., "An Investigation of Artifacts Resulting From Lossy Wavelet Compression of Digital Mammograms," 1995 AAPM, Works in Progress Poster, Boston, MA., 1995.

Pawlicki, T. and Gaborski, R.S., "Detection of Scatter Reducing Grids in Scanned Radiographic Images," 1995 AAPM, Works in Progress Poster, Boston, MA., 1995.

Gaborski, R. and Jang, B., "Enhancement of Computed Radiographic Images," IEEE Computer-Based Medical Systems, Lubbock, TX, 1995.

Jones, P., Daly, S., Gaborski, and R., Rabbani, M., "Comparative Study of Wavelet and DCT Decomposition with Equivalent Quantization and Encoding Strategies for Medical Images" SPIE Medical Imaging 1995, San Diego, California, 1995.

Jang, B., Gaborski, R., "Image Enhancement for Computed Radiographic Images," SPIE Medical Imaging 1995, San Diego, California, 1995.

Anderson, P., Gaborski,R. and Rao, A., " The Polynomial Classification System," Proceedings of the 47th Annula Conference of Society for Imaging Science and Technology, May, 1994, pp 508-510.

Barski, L., Gaborski, R.S. and Anderson, P., "A Neural Network Approach to the Segmentation of Digital Radiographic Images,?" ANNIE 93 Artificial Neural Networks in Engineering, St. Louis, MO, 1993

Rao, A., Anderson, P., and Gaborski, R., "A Hardware Polynomial Feature Net for OCR", IEE Third International Conference on Artificial Neural Networks, Brighton, UK, 1993.

Gaborski, R., Anderson, P., and Tilley, D., "Genetic Algorithm Selection for Handwritten Character Identification," International Conference on Neural Networks and Genetic Algorithms, Innsbruck, Austria, 1993.

Anderson, P. and Gaborski, R., "The Polynomial Method Augmented by Supervised training for Hand Printed Character recognition," Innsbruck, Austria, 1993.

Gaborski, R. and Barski, L., "Improved Readability of Machine Printed Characters," Advanced Technology Conference, US Postal Service, Washington, D.C., 1992.

Lee D., Srihari S., Gaborski R. . Baysian and Neural Network Pattern Recognition: a Theoretical Connection and Empirical Results with Handwritten Characters. In: Artificial Neural Networks and Statistical Pattern Recognition, 1991. Sethi I. K., Jain A. K. (Eds.) 89 - 108. Elsevier Science.

Gaborski, R. and Barski, L., "The Current State of cursive Script Recognition at Kodak," 87th InterPlant Conference- Electronic and Hybrid Imaging Systems, Rochester, NY, 1991.

Barski, L. and Gaborski, R., "Combined Segmentation and optical Character Recognition Using a Neural Network," 87th InterPlant Conference- Electronic and Hybrid Imaging Systems, Rochester, NY, 1991.

Gaborski, R.,"An Intelligent Character Recognition System Based on Neural Networks," Kodak Research Magazine, Spring, 1990.

Lee, D., Srihari, S., and Gaborski, R., "Experiments in Handwritten Character Recognition with Pattern Recognition and Neural Network Approaches," Advanced Technology Conference, US Postal Service, Washington, D.C., 1990.
 
 

Patents Issued ( inventor or co-inventor):

US06018590  Technique for Finding the Histogram Region of Interest Based on Landmark Detection for Improved Tonescale Reproduction of Digital Radiographic Images  01/26/2000

US05995682  Method for Resizing of a Digital Image 10/30/99

US05978518 Image Enhancement in Digital Image Processing 10/02/99

US05943435  Body Part Recognition in Radiographic Images 08/24/99

US05923775   Apparatus and Method for Signal Dependent Noise Estimation and Reduction in Digital Images    07/13/1999

US05862249   Automated Method and System for Determination of Positional Orientation of Digital Radiographic Images    01/19/1999

US05857030   Automated Method and System for Digital Image Processing of Radiologic Images Utilizing Artificial Neural Networks    01/05/1999

US05796862   Apparatus and Method for Identification of Tissue Regions in Digital Mammographic Images    08/18/1998

US05696805   Apparatus and Method for Identifying Specific Bone Regions in Digital X-ray Images    12/09/1997

US05661818   A Method and System for Detecting Grids in a Digital Image    08/26/1997

US05574803   Character Thinning using Emergent Behavior of Populations of Competitive Locallly Independent Processes    11/12/1996

US05553162   Method for Detecting Ink Jet or Dot Matrix Printing    09/03/1996

US05479523  Constructing Classification Weights Matrices for Pattern Recognition Systems using Reduced Element Feature Subsets    12/26/1995

US05442715   Method and Apparatus for Cursive Script Recognition    08/15/1995

US05426684  Technique for Finding the Histogram Region of Interest for Improved Tone Scale Reproduction of Digital Radiographic Images    06/20/1995

US05392447   Image-based Electronic Pocket Organizer with Integral Scanning Unit    02/21/1995

US05299269   Character Segmentation Using an Associative Memory for Optical Character Recognition    03/29/1994

US05212741   Pre-processing of Dot-Matrix/Ink Jet Printed Text for OCR    05/18/1993

US05054102   Self-centering Character Stroke Thickening for Optical Character Recognition    10/01/1991

US05052044   Correlated Masking Process for Deskewing, Filtering, and Recognition of Vertically Segmented Characters    09/24/1991

US05052043   Neural Network with Back Propagation Controlled through an Output Configdence Measure    09/24/1991

US05048097   Optical Character Recognition Neural Network System for Machine-printed Characters    09/10/1991

US04949392   Document Recognition with Automatic Indexing for Optical Character Recognition    08/14/1990

US04791679   Image Character Enhancement Using a Stroke Strengthening Kernel    12/13/1988