Exploiting Multimodal Synergy for Large Scale and Diverse Image Retrieval in Digital Archives

Pilot Research on Developing Theory and Applications of Community Generation

Sponsor: Air Force Research Laboratory

This is a three year effort of initiating a pilot research on developing the theory of community generation and its various applications in situation awareness. The ultimate goals of this effort include to advance the knowledge and the literature on Knowledge Discovery and Data Mining as well as Information Fusion for Situation Awareness and to promote the knowledge dissemination and the technology awareness resulted from the research to benefit the government and the whole nation through a suite of identified applications in situation awareness and the specific applications related to homeland security.

Based on the ultimate goals stated above, this project aims at the following two specific objectives: (1) to initiate a preliminary investigation on the theoretic foundation of the problems related to community generation; and (2) to study specific applications of community generation in situation awareness by identifying a suite of important applications in homeland security.

The technical merits of this project include the advancement of the knowledge and understanding in the literature not only in the data mining area but also across a wide spectrum of areas including information fusion, pattern recognition, and machine learning. The broader impacts of this project include a wide range of applications from homeland security to economic development to life science and engineering that will be benefited from the research of this project per se as well as the knowledge disseminated from this research.

Project Personnel:

PI:

Prof. Zhongfei (Mark) Zhang(SUNY Binghamton)

Students:

Bo Long (PhD student)
Jingzhou Hua (PhD student)
Ron Lewin (MS student)
Haricharan Kulkarni (MS Student)
Michael Decker (Undergraduate student)

AFRL Program Managers:

Dr. John Salerno
Dr. Barry McKinney

Collaborating Organizations:

IBM Watson Research Center (Dr. Philip Yu)
Google Research Labs (Dr. Xiaoyun Wu)
Air Force Research Laboratory (Dr. John Salerno)

Project Demonstrations:

A specific theory that we have developed (Link Discovery based on Correlation Analysis (LDCA)) is applied to solving for the money laundering crime investigation problem, for which a demo is developed on a real money laundering crime case provided by NIJ.

Publications:

Ruofei Zhang and Zhongfei (Mark) Zhang, Effective Image Retrieval Based on Hidden Concept Discovery in Image Database, IEEE Transaction on Image Processing, Volume 16, Number 2, February, 2006, pp 562 -- 572

Arif Ghafoor, Zhongfei (Mark) Zhang, Michael S. Lew, and Zhi-Hua Zhou, Guest Editors' Introduction to Machine Learning Approaches to Multimedia Information Retrieval, ACM Multimedia Systems Journal, Springer, 2006

Zhongfei (Mark) Zhang, Querying Non-Uniform Image Databases for Biometrics-Related Identification Applications, Sensor Review, Emerald Publishers, Volume 26, Number 2, April, 2006, pp 122-126

Ruofei Zhang and Zhongfei (Mark) Zhang, Empirical Bayesian Learning in the Relevance Feedback of Image Retrieval, Image and Vision Computing, Elsevier Science, Volume 24, Issue 3, March, 2006, pp 211-2233

Ruofei Zhang, Zhongfei (Mark) Zhang, Mingjing Li, Wei-Ying Ma, and Hong-Jiang Zhang, A Probabilistic Semantic Model for Image Annotation and Multi-Modal Image Retrieval, ACM Multimedia Systems Journal, the special issue of Using Machine Learning Approaches to Multimedia Information Retrieval, Springer, 2006

Jian Yao and Zhongfei (Mark) Zhang, Hierarchical Shadow Detection for Color Aerial Images, Computer Vision and Image Understanding, Elsevier Science, Volume 102, Issue 1, April, 2006, pp 60-699

Bo Long, Xiaoyun Wu, Zhongfei (Mark) Zhang, and Philip S. Yu, Unsupervised Learning on K-partite Graphs, Proc. ACM International Conference on Knowledge Discovery and Data Mining, ACM Press, Philadelphia, PA, USA, August, 2006[pdf]

Bo Long, Zhongfei (Mark) Zhang, Xiaoyun Wu, and Philip S. Yu, Spectral Clustering for Multi-Type Relational Data, Proc. International Conference on Machine Learning, ACM Press, Pittsburgh, PA, USA, June, 2006[pdf]

Xiao-Bing Xue, Zhi-Hua Zhou, and Zhongfei (Mark) Zhang, Improve Web Search Using Image Snippets, Proc. the 21st National Conference on Artificial Intelligence, AAAI Press, Boston, MA, USA, July, 2006

Jian Yao, Zhongfei (Mark) Zhang, Sameer Antani, Rodney Long, and George Thoma, Automatic Medical Image Annotation and Retrieval Using SEMI-SECC, Proc. IEEE International Conference on Multimedia and Expo, IEEE Computer Society Press, Toronto, Canada, July, 2006

Jian Yao, Sameer Antani, Rodney Long, and George Thoma, and Zhongfei (Mark) Zhang, Automatic Medical Image Annotation and Retrieval Using SECC, Proc. IEEE International Symposium on Computer Based Medical Systems, IEEE Computer Society Press, Salt Lake City, Utah, USA, June, 2006[pdf]

Bo Long, Zhongfei (Mark) Zhang, and Philip S. Yu, Combining Multiple Clusterings by Soft Correspondence, Proc. IEEE International Conference on Data Mining, IEEE Computer Society Press, New Orleans, LA, November, 2005[pdf]

Bo Long, Zhongfei (Mark) Zhang, and Philip S. Yu, Co-clustering by Block Value Decomposition, Proc. ACM International Conference on Knowledge Discovery and Data Mining, ACM Press, Chicago, IL, August, 2005[pdf]

Ruofei Zhang, Zhongfei (Mark) Zhang, and Sandeep Khanzode, A Data Mining Approach to Modeling Relationships Among Categories in Image Collection, proceedings of the 10th ACM International Conference on Knowledge Discovery and Data Mining (ACM KDD), Seattle, WA, Aug. 2004.[pdf]

Zhongfei (Mark) Zhang, John J. Salerno, Philip S. Yu, Jingzhou Hua, Ruofei Zhang, Maureen Regan, and Debra Cutler, Applying data mining in investigating money laundering crimes, ACM KDD 2003, Washington DC, Aug., 2003[pdf]

This material is based upon the work supported by the Air Force Research Lab under project No. F30602-02-M-V020 and FA8750-04-1-0234.

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the Air Force Research Lab.

Go back to the Multimedia Computing Research Lab homepage