• The KIPS Transactions:PartD
• /
• v.8D no.5
• /
• pp.507-512
• /
• 2001

Feature-based similarity has become an important research issue in multimedia database systems. The features of multimedia data are useful for discriminating between multimedia objects. the performance of conventional multidimensional data structures tends to deteriorate as the number of dimensions of feature vectors increase. The $R^{*}-Tree$ is the most successful variant of the R-Tree. In this paper, we propose a SOM-based $R^{*}-Tree$ as a new indexing method for high-dimensional feature vectors. The SOM-based $R^{*}-Tree$ combines SOM and $R^{*}-Tree$ to achieve search performance more scalable to high-dimensionalties. Self-Organizingf Maps (SOMs) provide mapping from high-dimensional feature vectors onto a two-dimensional space. The map is called a topological feature map, and preserves the mutual relationships (similarity) in the feature spaces of input data, clustering mutually similar feature vectors in neighboring nodes. Each node of the topological feature map holds a codebook vector. We experimentally compare the retrieval time cost of a SOM-based $R^{*}-Tree$ with of an SOM and $R^{*}-Tree$ using color feature vectors extracted from 40,000 images. The results show that the SOM-based $R^{*}-Tree$ outperform both the SOM and $R^{*}-Tree$ due to reduction of the number of nodes to build $R^{*}-Tree$ and retrieval time cost.

• Journal of the Institute of Convergence Signal Processing
• /
• v.4 no.2
• /
• pp.12-17
• /
• 2003

This paper proposes content-based image retrieval system with fuzzy ART neural network algorithm. Retrieving large database of image data, the clustering is essential for fast retrieval. However, it is difficult to cluster huge image data pertinently, Because current retrieval methods using similarities have several problems like low accuracy of retrieving and long retrieval time, a solution is necessary to complement these problems. This paper presents a content-based image retrieval system with neural network in order to reinforce abovementioned problems. The retrieval system using fuzzy ART algorithm normalizes color and texture as feature values of input data between 0 and 1, and then it runs after clustering the input data. The implemental result with 300 image data shows retrieval accuracy of approximately 87%.

• Journal of Information Processing Systems
• /
• v.14 no.1
• /
• pp.176-190
• /
• 2018

This paper presents a novel convolutional neural network based multi-feature fusion learning method for non-rigid 3D model retrieval, which can investigate the useful discriminative information of the heat kernel signature (HKS) descriptor and the wave kernel signature (WKS) descriptor. At first, we compute the 2D shape distributions of the two kinds of descriptors to represent the 3D model and use them as the input to the networks. Then we construct two convolutional neural networks for the HKS distribution and the WKS distribution separately, and use the multi-feature fusion layer to connect them. The fusion layer not only can exploit more discriminative characteristics of the two descriptors, but also can complement the correlated information between the two kinds of descriptors. Furthermore, to further improve the performance of the description ability, the cross-connected layer is built to combine the low-level features with high-level features. Extensive experiments have validated the effectiveness of the designed multi-feature fusion learning method.

• Journal of Institute of Control, Robotics and Systems
• /
• v.10 no.7
• /
• pp.597-603
• /
• 2004

In retrieving large database of image data, the clustering is essential for fast retrieval. However, it is difficult to cluster a number of image data adequately. Moreover, current retrieval methods using similarities are uncertain of retrieval accuracy and take much retrieving time. In this paper, a suggested image retrieval system combines Fuzzy ART neural network algorithm to reinforce defects and to support them efficiently. This image retrieval system takes color and texture as specific feature required in retrieval system and normalizes each of them. We adapt Fuzzy ART algorithm as neural network which receive normalized input-vector and propose improved Fuzzy ART algorithm. The result of implementation with 200 image data shows approximately retrieval ratio of 83%.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.6 no.2
• /
• pp.95-106
• /
• 2007

This paper proposes an effective tempo feature extraction method for music information retrieval. The tempo information is modeled by the narrow-band temporal modulation components, which are decomposed into a modulation spectrum via joint frequency analysis. In implementation, the tempo feature is directly extracted from the modified discrete cosine transform coefficients, which is the output of partial MP3(MPEG 1 Layer 3) decoder. Then, different features are extracted from the amplitudes of modulation spectrum and applied to different music information retrieval tasks. The logarithmic scale modulation frequency coefficients are employed in automatic music emotion classification and music genre classification. The classification precision in both systems is improved significantly. The bit vectors derived from adaptive modulation spectrum is used in audio fingerprinting task That is proved to be able to achieve high robustness in this application. The experimental results in these tasks validate the effectiveness of the proposed tempo feature.

• ETRI Journal
• /
• v.29 no.5
• /
• pp.700-702
• /
• 2007

We present a relevance feedback approach based on multi-class support vector machine (SVM) learning and cluster-merging which can significantly improve the retrieval performance in region-based image retrieval. Semantically relevant images may exhibit various visual characteristics and may be scattered in several classes in the feature space due to the semantic gap between low-level features and high-level semantics in the user's mind. To find the semantic classes through relevance feedback, the proposed method reduces the burden of completely re-clustering the classes at iterations and classifies multiple classes. Experimental results show that the proposed method is more effective and efficient than the two-class SVM and multi-class relevance feedback methods.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.10C
• /
• pp.1444-1450
• /
• 2004

MPEG-7 based on XML represents various information of multimedia data's contents. and it support search and browsing by user's wants. But, MPEG-7 standard don't support retrieval method and Many XML Indexing is not compatible to retrieval MPEG-7 documents. So Much research activity and interest has emerged recently in retrieval MPEG-7 documents. In our paper, we suppose a hierarchical index based on MPEG-7 document's structural information, and review how to query processing based on high level feature description.

• Journal of Information Processing Systems
• /
• v.17 no.5
• /
• pp.1020-1033
• /
• 2021

Nowadays, online or mobile social network services (SNS) are very popular and widely spread in our society and daily lives to instantly share, disseminate, and search information. In particular, SNS such as YouTube, Flickr, Facebook, and Amazon allow users to upload billions of images or videos and also provide a number of multimedia information to users. Information retrieval in multimedia-rich SNS is very useful but challenging task. Content-based media retrieval (CBMR) is the process of obtaining the relevant image or video objects for a given query from a collection of information sources. However, CBMR suffers from the dimensionality curse due to inherent high dimensionality features of media data. This paper investigates the effectiveness of the kernel trick in CBMR, specifically, the kernel principal component analysis (KPCA) for dimensionality reduction. KPCA is a nonlinear extension of linear principal component analysis (LPCA) to discovering nonlinear embeddings using the kernel trick. The fundamental idea of KPCA is mapping the input data into a highdimensional feature space through a nonlinear kernel function and then computing the principal components on that mapped space. This paper investigates the potential of KPCA in CBMR for feature extraction or dimensionality reduction. Using the Gaussian kernel in our experiments, we compute the principal components of an image dataset in the transformed space and then we use them as new feature dimensions for the image dataset. Moreover, KPCA can be applied to other many domains including CBMR, where LPCA has been used to extract features and where the nonlinear extension would be effective. Our results from extensive experiments demonstrate that the potential of KPCA is very encouraging compared with LPCA in CBMR.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.59 no.5
• /
• pp.990-995
• /
• 2010

Harris corner detector is commonly used to detect feature points for recognizing 2-D or 3-D objects. However, the feature points calculated from both of query and target objects need to be same positions to guarantee accurate recognitions. In order to check the positions of calculated feature points, we generate a Huffman tree which is based on adjacent feature values as inputs. However, the structures of two Huffman trees will be same as long as both of a query and targets have same feature values no matter how different their positions are. In this paper, we sort feature values and calculate the Euclidean distances of coordinates between two adjacent feature values. The Huffman Tree is generated with these Euclidean distances. As a result, the information of point locations can be included in the generated Huffman tree. This is the main strategy for accurate recognitions. We call this system as the HRM(Hybrid Retrieval Machine). This system works very well even when artificial random noises are added to original data. HRM can be used to recognize biological data such as proteins, and it will curtail the costs which are required to biological experiments.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.8 no.3
• /
• pp.122-130
• /
• 2009

The paper presents an algorithm which uses code vectors to represent comer geometry information for searching the similar images from a database. The comers have been extracted by finding the intersections of the detected lines found using Hough transform. Taking the comer as the center coordinate, the angles of the intersecting lines are determined and are represented using code vectors. A code book has been used to code each comer geometry information and indexes to the code book are generated. For similarity measurement, the histogram of the code book indexes is used. This result in a significant small size feature matrix compared to the algorithms using color features. Experimental results show that use of code vectors is computationally efficient in similarity measurement and the comers being noise invariant produce good results in noisy environments.