Summary

The Columbia TAG (Transductive Annotation by Graph) system is deigned to facilitate rapid retrieval and exploration of large image and video collections. It incorporates novel graph-based label propagation methods and intuitive graphic user interfaces (GUI) that allow users quickly browse and annotate a small number of images/videos, and then in real or near real time receive refined labels for all remaining unlabeled data in the collection. Using such refined labels, additional positive images/videos matching user's interest can be quickly discovered. It can be used as a fast search system alone, or a bootstrapping system for developing additional target recognition tools needed in critical image application domains such as intelligence, surveillance, consumer, biomedical, and Web. The diagram of TAG system is shown below.

The system diagram and usage modes of Columbia TAG (Transductive Annotation by Graph)System.

Video Demo (download high quality WMV movie)

People

Jun Wang and Shih-Fu Chang

Publications

1. Jun Wang, Shih-Fu Chang. Columbia TAG System - Transductive Annotation by Graph Version 1.0. ADVENT Technical Report #225-2008-3 Columbia University, October 2008. [pdf]

2. Jun Wang, Yu-Gang, and Shih-Fu Chang, Label Diagnosis through Self Tuning for Web Image Search. IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR), Miami, Florida, June 20-25, 2009. [pdf][poster]

3. Tony Jebara, Jun Wang, and Shih-Fu Chang, Graph Construction and b-Matching for Semi-Supervised Learning. International Conference on Machine Learning (ICML), Montreal, Quebec, June 14-18, 2009. [pdf][ppt][poster][video][code]

4. Jun Wang, Tony Jebara, and Shih-Fu Chang, Graph Transduction via Alternating Minimization. International Conference on Machine Learning (ICML), Helsinki, Finland, July 5-9, 2008. [pdf][ppt][poster][video][code]

5. Jun Wang, Shih-Fu Chang, Xiaobo Zhou, Stephen T. C. Wong. Active Microscopic Cellular Image Annotation by Superposable Graph Transduction with Imbalanced Labels. In IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR), Anchorage, Alaska, USA, June 24-26 2008. [pdf][poster]

For problems or questions regarding this web site contact The Web Master.
Last updated: May. 11, 2009.