累積由過往之查詢期間(Query Session)所學習到的知識用以增進以內容為基礎的影像資料庫檢索 (Content-Based Image Retrieval)之檢索效果是目前影像檢索研究的主要課題之一。經由查詢期間歷史(Query Session History)所得到的資訊可以有效的克服語意隔閡 (Semantic Gap)並顯著的增進檢索效果與效率。本篇論文提出一個能以隱含語意空間(Hidden Semantic Space)累計過去查詢經驗的影像檢索系統。首先，在每次的查詢期間根據使用者提供之相關回饋(Relevance Feedback)學習出一個支援向量分類器，並以該分類器估算影像含有該查詢期間所代表之隱含語意的機率。以多次查詢期間所習得之隱含語意機率建構出隱含語意空間，在往後的查詢期間便可用此隱含語意空間所提供之隱含語意特徵進行檢索。為了避免隱含語意空間因持續累積查詢期間的資訊而造成空間維度過大，我們提出以圖像為基礎之群聚整合學習的方法根據現有的隱含語意特徵對影像進行分群以縮減隱含語意空間的大小。分群後之每個群聚即為一個整合過之隱含語意概念。經由設計的實驗，我們可證實我們所提出的以圖像為基礎之群聚整合學習方法能有效率的降低隱含語意之維度，並保持隱含語意空間的有效性及可靠性且提升檢索系統的效能表現。

Inter-session learning in content-base image retrieval (CBIR) makes user take advantage of the information learned from previous query session. Many works have been proposed for the inter-session learning. In this thesis, the basis is a framework using a hidden semantic space to accumulate the inter-session information. At first, we use the SVM classifiers trained in short-term learning to initialize the hidden semantic space with a probabilistic model. To maintain the hidden semantic space in a proper size, we propose a novel framework based on graph-based cluster ensemble. Each time the hidden semantic space is over-expanding, we use our proposed dimension reducing method to construct a compact, effective, meaningful, and lower dimensional hidden semantic space. With the hidden semantic space, a long-term learning scheme is performed. Our experimental results demonstrate that the graph-based cluster ensemble scheme works well and efficient in our long-term learning CBIR system. The scheme takes short time but provides stable and reliable results.

中文摘要 i
Abstract ii

1. INTRODUCTION 1
2. RELATED WORK 5
2.1. Short-term Learning 5
2.2. Long-term Learning 6
2.2.1. Extended Relevance Feedback 7
2.2.2. Semantic Similarity Learning 8
2.2.3. Semantic Space Construction 10
2.3. Cluster Ensemble 11
3. MOTIVATION 18
4. PROPOSED METHOD 20
4.1. Short-term Learning 20
4.1.1. Feature Extraction 20
4.1.2. Image Similarity Evaluation 21
4.1.3. Learning from Relevance Feedback 22
4.2. Hidden Semantic Space Initialization 28
4.3. Reducing the Semantic Space Dimension 30
4.3.1. Hidden Semantic Space and Cluster Ensemble 31
4.3.2. Cluster-based Similarity Partition Algorithm (CSPA) 32
4.3.3. Meta-Clustering Algorithm (MCLA) 32
4.3.4. Multilevel Graph Partition (METIS) 34
4.3.5. One-against-others SVM Classifier 38
4.4. Long-term Learning 39
5. EXPERIMENTAL RESULTS 45
5.1. Noise-free Relevance Feedbacks 47
5.2. Noisy Relevance Feedbacks 49
5.3. Discussion 50
6. CONCLUSION 61
7. REFERENCES 63

[1] Y. Rui, T. S. Huang, M. Ortega, and S. Mehrotra, “Relevance Feedback: A Power Tool for Interactive Content-Based Image Retrieval,” IEEE Trans. Circuits and Systems for Video Technology, vol. 8, no. 5, pp. 644-655, 1998.
[2] Y. Ishikawa, R. Subramanya, and C. Faloutsos, “MindReader: Query Databases through multiple examples,” Proc. 24th Int. Conf. Very Large Databases, pp.218-227, 1998.
[3] G. Giacinto and F. Roli, ”Bayesian Relevance Feedback for Content-Based Image Retrieval,” Pattern Recognition, vol.37, issue 7, pp. 1499-1508, 2004.
[4] I. J. Cox, M. L. Miller, T. P. Minka, T. V. Papathomas, and P. N. Yianilos, “The Bayesian Image Retrieval System, PicHunter: Theory, Implementation, and Psychophysical Experiments,” IEEE Trans. Image Processing, vol. 9, pp. 20-37, 2000.
[5] R. Zhang and Z. Zhang, “Stretching Bayesian Learning in the Relevance Feedback of Image Retrieval,” EECV, vol. 3, pp.355-367, 2004.
[6] I. Bartolini, P. Ciaacai, and F. Waas, “FeedbackBypass: A New Approach to Interactive Similarity Query Processing,” Proc. 27th Int. Conf. Very Large Databases, pp.201-210, 2001.
[7] X. Zhou, Q. Zhang, L. Liu, L. Zhang, and B. Shi, “An image retrieval method based on analysis of feedback sequence log,” Pattern Recognition Letters, vol.24, Issue 14, pp. 2499-2508, 2003.
[8] J. Han, K. N. Ngan, M. Li, and H. Zhang, “Learning Semantic Concepts form User Feedback Log for Image Retrieval,” Proc. ICME, 2004.
[9] J. Fournier and M. Cord, “Long-term Similarity Learning in Content-based Image Retrieval,” Proc. ICIP, 2002.
[10] H. Y. Bang, C. Zhang and T. Chen, “Semantic Propagation from Relevance Feedbacks,” Proc. ICME, 2004.
[11] P. H. Gosselin and M. Cord, ”Semantic Kernel Learning For Interactive Image Retrieval,” Proc. ICIP, 2005.
[12] X. He, O. King, W. Y. Ma, M. Li, and H. J. Zhang, “Learning a Semantic Space form User’s Relevance Feedback for Image Retrieval,” IEEE Trans. Circuit and Systems for Video Technology, vol. 13, no.1, pp. 39-48, 2003.
[13] J. W. Tung and C. T. Hsu, “Learning Hidden Semantic Cues Using Support Vector Clustering,” Proc. ICIP, 2005.
[14] A. L. N. Fred and A. K. Jain, “Data Clustering Using Evidence Accumulation,” Proc. ICPR, 2002.
[15] A. Topchy, A. K. Jain, W. Punch, “Clustering Ensembles: Models of Consensus and Weak Partitions,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 27, no.12, pp. 1866-1881, 2005.
[16] A. Strehl and J. Ghosh, “Cluster Ensembles – A Knowledge Reuse Framework for Combining Multiple Partitions,” Journal of Machine Learning Research 3, pp. 583-617, 2002.
[17] G. Karypis and V. Kumar, “A fast and high quality multilevel scheme for partitioning irregular graphs,” SIAM Journal on Scientific Computing, vol. 20, no. 1, pp. 359-392, 1999.
[18] G. Karypis, R. Aggarwal, V. Kumar, and S. Shekhar, “Multilevel Hypergraph Partitioing: Applications in VLSI domain,” IEEE Trans. VLSI systems, vol. 7, no.1, pp. 96-79, 1999.
[19] D. Comaniciu and P. Meer, “Mean Shift: A Robust Approach toward Feature Space Analysis,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 23, no.5, pp. 603-619, 2002.
[20] L. G. Shapiro and G. C. Stockman, Computer Vision, Prentice Hall, 2001.
[21] J. Li, J. Z. Wang, and G. Wiederhold, “IRM: integrated region matching for image retrieval,” Proc. 8th ACM intl. Conf. Multimedia, 2000.
[22] H. Greenspan, J. Goldberger and Lenny Ridel, “A Continuous Probabilistic Framework for Image Matching,” Journal of Computer Vision and Image Understanding, 1-23, pp. 96-129, 1998.
[23] N. Cristianini and J. Shawe-Taylor, An Introduction To Support Vector Machines, Cambridge University Press, 2000.
[24] J. C. Platt, “Sequential Minimal Optimization: A Fast Algorithm for Train Support Vector Machines,” Tech. Report MSR-TR-98-14, Microsoft Research, 1998.
[25] J. C. Platt, “Probabilistic Outputs for Support Vector Machines and Comparisons to Regularized Likelihood Methods,” Advances in Large Margin Classifiers, MIT Press, pp. 61-74, 2000.
[26] G. Karypis and V. Kumar, “Analysis of Multilevel Graph Partitioning,” Technical Report TR 95-037, Department of Computer Science, University of Minnesota, 1995.
[27] C. M. Fiduccia and R. M. Mattheyses, “A Linear-Time Heuristic for Improving Network Partitions,” Proc.19th Conf. IEEE Design Automation, pp. 175-181, 1982.