基於社群標籤之視覺化圖片查詢意識

Visualizing Image Query Sense by Social Tags

戴瑋彥(Wei-Yen Day)

指導教授 ： 鄭卜壬

國立臺灣大學/電機資訊學院/資訊工程學研究所/碩士(2010年)

https://doi.org/10.6342/NTU.2010.01396

摘要

本論文提出了一個嶄新的圖片檢索概念：以視覺化概念(visual concepts) 來輔助圖片搜尋。視覺化概念往往可以輔助原先之圖片查詢，並進一步搜尋得到更視覺化效果以及更有明確主題和富有使用者觀感意識之圖片。過去有許多文獻對於概念輔助圖片搜尋有進一步研究，但視覺化概念輔助圖片檢所之問題並未被仔細討論。因此，本論文根據視覺化、主題性、使用者觀感意識等三種角度提出了不同的對應方法來對圖片查詢之視覺化概念做排序。有鑑於社群標籤(social tag)近年來被廣泛運用，以及其對圖片之精確性及描述性可為之利用，本論文之視覺化概念來源皆來自於社群標籤，除了探討社群標籤之困難度及複雜程度，並依照提出之方法做不同角度之分析與排序，以進一步對使用者之圖片查詢做查詢擴張(query expansion)，來達到增加圖片檢索效能之目的。本論文之實驗應用在真實世界之資料，藉由使用者行為探討(user study) 來達到評估方法效能之目的。實驗結果顯示視覺化概念的確幫助提升圖片檢索之效能，並可進一步藉由視覺化概念來增加圖片檢索之視覺化程度、主題性、富有使用者觀感，以及圖片檢索結果之涵蓋廣泛程度。

關鍵字

圖片搜尋 ； 圖片查詢 ； 圖片查詢意識 ； 社群標籤

並列摘要

In this paper, we present an approach for visualizing image query senses. Image queries usually have several senses, which can describe the meanings of themselves. However, senses like ‘hot’ might not be concrete, thus we need to find out visual concepts to visualize these image query senses. We propose a novel approach to discover the visual concepts for image queries based on several statistical scores and social tags from Flickr, and further help improve image search by visualizing their senses. To evaluate the effectiveness of our approach, we test the found concepts on real world queries and images. Both the experimental results, conducted for image retrieval and concepts evaluation, demonstrate that the approach can substantially improve the traditional image search engine, which retrieve only relevant images, and show that the visual concepts for image query senses can be utilized to enhance the effectiveness of image retrieval.

並列關鍵字

Image search ； image query ； image query sense ； social tags

參考文獻

[1] M. Arevalillo-Herr’aez, F. J. Ferri, and J. Domingo. A naive relevance feedback model for content-based image retrieval using multiple similarity measures. Pattern Recogn., 43(3):619–629, 2010.

[4] Y.-C. Chang and H.-H. Chen. Using an image-text parallel corpus and the web for query expansion in cross-language image retrieval. pages 504–511, 2008.

[6] C.-C. Chiang, J.-Y. Wu, M.-T. Yang, and W.-K. Tai. Independent query refinement and feature re-weighting using positive and negative examples for content-based image retrieval. Multimedia Tools Appl., 41(1):27–53, 2009.

NY, USA, 2005. ACM.

[8] J. Deng,W. Dong, R. Socher, L. jia Li, K. Li, and L. Fei-fei. Imagenet: A large-scale hierarchical image database. In In CVPR, 2009.