Video Data Mining
Data mining techniques that are successful in text and transaction data
may not simply apply to image data that are non-structured. It is not a
trivial task to discover meaningful visual patterns in image databases,
because the spatial dependency and content variations in the visual data
greatly challenge most existing methods.
We propose a novel and effective multilevel approach to cope with these
difficulties for mining spatial co-location visual patterns. Specifically,
the novelty of this work lies in the following components: (1) an efficient
pattern discovery and summarization method that handles spatial dependency;
and (2) an effective multilevel probabilistic method to tame the content
uncertainties of visual patterns. This new approach learns a hierarchical
generative model of the image database in an unsupervised fashion, and
can be applied to various applications.
Demo Sequences of On-line Data Mining
[These videos are for the purpose of demonstrating the technology only. Any reproduction or propagation of these video is
explicitly
forbidden and the privacy of the persons in the video should be protected to the full extent.]
|
|
|
A
dance group
|
A boy in a pool
|
A mom with the
daughter
|
|
|
Note: [top-left] results of on-line video data mining that identifies video context of the target (e.g., the
headin this case). [top-right] robust information integration that combines all the predictions from the video context.
|
A boy in a room
|
A girl and her dady in
woods |
Note: [bottom-left] comparison with a dedicated head tracker. [bottom-right] our result where the yellow bounding box highlights
the output the tracked target. |
The major subject of each sequence is a small kid, and the tracking of the head of the kid is of great interest. These
sequences exhibist many challenges such as occlusion, background clutter, appearance changes, etc. In addition, these
video are recorded by amateurs with a hand-held camera. The tracking task here is very difficult. We developed a method
that introduces on-line data mining to visual tracking, which automatically discovers the visual context of the target and
the use of the context make the tracking quite stable and robust (for details please see our papers).
Publication:
-
Ming Yang, Gang Hua and Ying Wu, "Context-Aware Visual Tracking", IEEE Trans. on
Pattern Analysis and Machine Intelligence, 2008
-
Ming Yang, Ying Wu and Shihong Lao,
"Intelligent Collaborative Tracking by Mining Auxiliary Objects",
in
Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR
06), New York City, NY, June 17-22, 2006. [PDF]
-
Junsong Yuan, Ying Wu and Ming Yang, "Discovery of Collocation patterns: from Visual
Words
to Visual Phrases", in Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR'07), Minneapolis, MN, June 2007
-
Junsong Yuan, Ming Yang and Ying Wu, "From Frequent Itemsets To Semantically Meaningful Visual
Patterns", in Proc. ACM Int'l Conf. on Knowledge Discovery and Data Mining (SIGKDD'07), San Jose, CA, August 2007
-
Ming Yang, Ying Wu and Shihong Lao, "Mining Auxiliary Objects for Tracking by Multibody
Grouping", in Proc. IEEE Int'l Conf. on Image Processing (ICIP'07), San Antonio, TX, Sept. 2007
-
Junsong Yuan, Zhu Li, Yun Fu, Ying Wu and Thomas S. Huang, "Common Spatial Pattern Discovery by
Efficient Candidate Pruning", in Proc. IEEE Int'l Conf. on Image Processing (ICIP'07), San Antonio, TX, Sept. 2007
-
Fan Jiang, Ying Wu and Aggelos K. Katsaggelos, "Abnormal Event Detection From Surveillance
Video By Dynamic Hierarchical Clustering", in Proc. IEEE Int'l Conf. on Image Processing (ICIP'07), San Antonio, TX, Sept. 2007
Back to Video Research
Updated 07/2006. Copyright ©
2003-2006 Ying
Wu