ABSTRACT
This work represents an initial investigation into determining whether correlations actually exist between metadata and content descriptors in multimedia datasets. We provide a quantitative method for evaluating whether the hue of images on the WWW is correlated with the occurrence of color-words in metadata such as URLs, image names, and attendant text. It turns out that such a correlation does exist: the likelihood that a particular color appears in an image whose URL, name, and/or attendant text contains the corresponding color-word is generally at least twice the likelihood that the color appears in a randomly chosen image on the WWW. While this finding might not be significant in and of itself, it represents an initial step towards quantitatively establishing that other, perhaps more useful correlations exist. These correlations form the basis for exciting novel approaches that leverage semi-supervised datasets, such as the WWW, to overcome the semantic gap that has hampered progress in multimedia information retrieval for some time now.
- Categorical image search demo. http://vision.ece.ucsb.edu/multimedia/search.html.Google Scholar
- Cortina: Large-scale, content-based image retrieval on the www.http://vision.ece.ucsb.edu/multimedia/cortina.html.Google Scholar
- The DMOZ open directory project. http://www.dmoz.org.Google Scholar
- Google image search. http://www.google.com.Google Scholar
- Yahoo image search. http://www.yahoo.com.Google Scholar
- K. Barnard and D. Forsyth. Clustering art. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 434--439, 2001.Google ScholarCross Ref
- K. Barnard and D. Forsyth. Learning the semantics of words and pictures. In Proceedings of the IEEE International Conference on Computer Vision, pages 408--415, 2001.Google ScholarCross Ref
- S. Newsam, B. Sumengen, and B. S. Manjunath. Category-based image retrieval. In Proceedings of the IEEE International Conference on Image Processing, volume 3, pages 596--599, 2001.Google ScholarCross Ref
- T. Quack. Cortina: A system for large-scale, content-base web image retrieval and the semantics within. Master's thesis, Swiss Federal Institute of Technology Zurich, April 2004. Google ScholarDigital Library
- S. Sclaroff, M. L. Cascia, S. Sethi, and L. Taycher. Unifying textual and visual cues for content-based image retrieval on the world wide web. Computer Vision and Image Understanding (CVIU), 75(1):86--98, 1999. Google ScholarDigital Library
Index Terms
- Seeing and reading red: hue and color-word correlation in images and attendant text on the WWW
Recommendations
Quality evaluation of tourmaline red based on uniform color space
Based on the uniform color space $$\hbox {CIE}\,1976\hbox {L}^{{*}}\hbox {a}^{{*}}\hbox {b}^{{*}},$$CIE1976Lźaźbź, the red colors of 310 tourmalines were conducted with the research of color colorimetry. In terms of the quantitative analysis of color ...
Flexible color contrast enhancement method for red-green deficiency
AbstractColor images are widely used to disseminate information via websites and smartphone applications. Red-green deficients may have difficulty distinguishing the colors, and this would cause ineffective visual communication. Recently, a naturalness ...
Transductive Multilabel Learning via Label Set Propagation
The problem of multilabel classification has attracted great interest in the last decade, where each instance can be assigned with a set of multiple class labels simultaneously. It has a wide variety of real-world applications, e.g., automatic image ...
Comments