Complementary relevance feedback-based content-based image retrieval
DOI 10.1007/s11042-013-1693-4
Complementary relevance feedback-based content-based image retrieval
Zhongmiao Xiao & Xiaojun Qi
# Springer Science+Business Media New York 2013
Abstract We propose a complementary relevance feedback-based content-based image retrieval (CBIR) system. This system exploits the synergism between short-term and long- term learning techniques to improve the retrieval performance. Specifically, we construct an adaptive semantic repository in long-term learning to store retrieval patterns of historical query sessions. We then extract high-level semantic features from the semantic repository and seamlessly integrate low-level visual features and high-level semantic features in short- term learning to effectively represent the query in a single retrieval session. The high-level semantic features are dynamically updated based on users’ query concept and therefore represent the image’s semantic concept more accurately. Our extensive experimental results demonstrate that the proposed system outperforms its seven state-of-the-art peer systems in terms of retrieval precision and storage space on a large scale imagery database.
Keywords Content-based image retrieval . Relevance feedback model . Semantic features . Long-term learning
1 Introduction
With the amount of digital photography data growing at an accelerating rate, the develop- ment of efficient image retrieval systems to find images of interest in this haystack of data has become an active research area in recent years [19]. Content-based image retrieval (CBIR) has emerged as one of the solutions to overcome the limitations entailed by text- based image retrieval and has evolved significantly since the early 1990s. It allows users to directly submit image examples, object sketches, or other low-level visual information (e.g., color, texture, and shape features) to find images of interest by using image processing and similarity matching
Complementary relevance feedback-based content-based image retrieval
Zhongmiao Xiao & Xiaojun Qi
# Springer Science+Business Media New York 2013
Abstract We propose a complementary relevance feedback-based content-based image retrieval (CBIR) system. This system exploits the synergism between short-term and long- term learning techniques to improve the retrieval performance. Specifically, we construct an adaptive semantic repository in long-term learning to store retrieval patterns of historical query sessions. We then extract high-level semantic features from the semantic repository and seamlessly integrate low-level visual features and high-level semantic features in short- term learning to effectively represent the query in a single retrieval session. The high-level semantic features are dynamically updated based on users’ query concept and therefore represent the image’s semantic concept more accurately. Our extensive experimental results demonstrate that the proposed system outperforms its seven state-of-the-art peer systems in terms of retrieval precision and storage space on a large scale imagery database.
Keywords Content-based image retrieval . Relevance feedback model . Semantic features . Long-term learning
1 Introduction
With the amount of digital photography data growing at an accelerating rate, the develop- ment of efficient image retrieval systems to find images of interest in this haystack of data has become an active research area in recent years [19]. Content-based image retrieval (CBIR) has emerged as one of the solutions to overcome the limitations entailed by text- based image retrieval and has evolved significantly since the early 1990s. It allows users to directly submit image examples, object sketches, or other low-level visual information (e.g., color, texture, and shape features) to find images of interest by using image processing and similarity matching
References: 1. Aksoy S, Haralick RM (2001) Feature normalization and likelihood-based similarity measures for image retrieval. Pattern Recogn Lett Spec Issue Image Video Indexing Retr 22(5):563–582 2 Image Processing, Brussels, Belgium, pp. 2473–2476 3 4. Chen Y, Rege M, Dong M, Fotouhi F (2007) Deriving semantics for image clustering from accumulated user feedbacks. In: 15th ACM Int. Conf. on Multimedia, Augsburg, Germany, pp. 313–316 5 8. Douze M, Jegou H, Sandhawalia H, Amsaleg L, Schmid C (2009) Evaluation of GIST descriptors for web-scale image search. In: Proc of the ACM Int Conf on Image and Video Retrieval, Santorini, Fira, Greece, Article No 9. Fechser S, Chang R, Qi X (2010) Inter-query semantic learning approach to image retrieval. In: IEEE Int Conf on Acoustics, Speech, and Signal Processing, Dallas, Texas, USA, pp 10. Han J, Ngan KN, Li M, Zhang HJ (2005) A memory learning framework for effective image retrieval. 12. Hoi SCH, Lyu MR, Jin R (2006) A unified log-based relevance feedback scheme for image retrieval. 14. Liu Y, Chen X, Zhang C, Sprague A (2009) Semantic clustering for region-based image retrieval. J Vis Comm Image Represent 20(2):157–166 15. Oliva A, Torralba A (2001) Modeling the shape of the scene: a holistic representation of the spatial envelope. Int J Comput Vis 42(3):145–175 16 18. Shah-Hosseini A, Knapp GM (2006) Semantic image retrieval based on probabilistic latent semantic analysis. In: 14th ACM Int Conf on Multimedia, Santa Barbara, CA, USA, pp. 703–706 19 21. Yin PY, Bhanu B, Chang KC, Dong A (2008) Long-term cross-session relevance feedback using virtual features 22. Yoshizawa T, Schweitzer H (2004) Long-term learning of semantic grouping from relevance feedback. In: 6th ACM SIGMM Int Workshop on Multimedia Information Retrieval, New York, NY, USA, pp