Image Clustering Using Multimodal Keywords
نویسندگان
چکیده
•Extending our previous work on visual keywords, we use the concept of template-based visual keywords using MPEG-7 color descriptors. MPEG-7, also called the Multimedia Content Description Interface, has been a standard for many years. These color descriptors have the ability to characterize perceptual color similarity and need relatively low complexity operations to extract them, besides being scalable and interoperable. We then demonstrate the power of these visual keywords for image clustering, when used in tandem with textual keyword annotations, in the context of latent semantic analysis, a popular technique in classical information retrieval which has been used to reveal the underlying semantic structure of document collections.
منابع مشابه
Discriminating Image Senses by Clustering with Multimodal Features
We discuss Image Sense Discrimination (ISD), and apply a method based on spectral clustering, using multimodal features from the image and text of the embedding web page. We evaluate our method on a new data set of annotated web images, retrieved with ambiguous query terms. Experiments investigate different levels of sense granularity, as well as the impact of text and image features, and globa...
متن کاملMultimodal medical image fusion based on Yager’s intuitionistic fuzzy sets
The objective of image fusion for medical images is to combine multiple images obtained from various sources into a single image suitable for better diagnosis. Most of the state-of-the-art image fusing technique is based on nonfuzzy sets, and the fused image so obtained lags with complementary information. Intuitionistic fuzzy sets (IFS) are determined to be more suitable for civilian, and medi...
متن کاملCross-Modal Image Clustering via Canonical Correlation Analysis
A new algorithm via Canonical Correlation Analysis (CCA) is developed in this paper to support more effective crossmodal image clustering for large-scale annotated image collections. It can be treated as a bi-media multimodal mapping problem and modeled as a correlation distribution over multimodal feature representations. It integrates the multimodal feature generation with the Locality Linear...
متن کاملRegistration-Based Segmentation Using the Information Bottleneck Method
We present two new clustering algorithms for medical image segmentation based on the multimodal image registration and the information bottleneck method. In these algorithms, the histogram bins of two registered multimodal 3D-images are clustered by minimizing the loss of mutual information between them. Thus, the clustering of histogram bins is driven by the preservation of the shared informat...
متن کاملمرور مؤثر نتایج جستجوی تصاویر با تلخیص بصری و متنوع از طریق خوشهبندی
With unprecedented growth in production of digital images and use of multimedia references, requirement of image and subject search has been increased. Systematic processing of this information is a basic prerequisite for effective analysis, organization and management of it. Likewise, large collections of images have been made available on the Web and many search engines have provided the poss...
متن کامل