• The Journal of the Acoustical Society of Korea
• /
• v.24 no.8
• /
• pp.498-504
• /
• 2005

We introduce a Korean speech recognition platform (ECHOS) developed for education and research Purposes. ECHOS lowers the entry barrier to speech recognition research and can be used as a reference engine by providing elementary speech recognition modules. It has an easy simple object-oriented architecture, implemented in the C++ language with the standard template library. The input of the ECHOS is digital speech data sampled at 8 or 16 kHz. Its output is the 1-best recognition result. N-best recognition results, and a word graph. The recognition engine is composed of MFCC/PLP feature extraction, HMM-based acoustic modeling, n-gram language modeling, finite state network (FSN)- and lexical tree-based search algorithms. It can handle various tasks from isolated word recognition to large vocabulary continuous speech recognition. We compare the performance of ECHOS and hidden Markov model toolkit (HTK) for validation. In an FSN-based task. ECHOS shows similar word accuracy while the recognition time is doubled because of object-oriented implementation. For a 8000-word continuous speech recognition task, using the lexical tree search algorithm different from the algorithm used in HTK, it increases the word error rate by $40\%$ relatively but reduces the recognition time to half.

• Industrial Engineering and Management Systems
• /
• v.8 no.4
• /
• pp.239-246
• /
• 2009

This paper deals with the makespan minimization problem of job shops. The problem is known as one of hard problems to optimize, and therefore, many heuristic methods have been proposed by many researchers. The aim of this study is also to propose a heuristic scheduling method for the problem. However, the difference between the proposed method and many other heuristics is that the proposed method is based on depth-first branch and bound, and thus it is possible to find an optimal solution at least in principle. To accelerate the search, when a node is judged hopeless in the search tree, the proposed flexible branch and bound method can indicate a higher backtracking node. The unexplored nodes are stored and may be explored later to realize the strict optimization. Two methods are proposed to generate the backtracking point based on the critical path of the current best feasible schedule, and the minimum lower bound for the makespan in the unexplored sub-problems. Schedules are generated based on Giffler and Thompson's active schedule generation algorithm. Acceleration of the search by the flexible branch and bound is confirmed by numerical experiment.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2007.11a
• /
• pp.459-463
• /
• 2007

Overlay multicast is an emerging technology for next generation Internet service to various groups of multicast members. It will take the place of traditional IP multicast which is not widely deployed due to the complex nature of its technology. The overlay multicast which effectively reduces processing at IP routers can be easily deployed on top of a densely connected IP network. An end-to-end delay problem is considered which is serious in the multicast service. To periodically optimize the route in the overlay network and to minimize the maximum end-to-end delay, overlay multicast tree is investigated with genetic Algorithm. Outstanding experimental results are obtained which is comparable to the optimal solution and the tabu search.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.31A no.7
• /
• pp.55-62
• /
• 1994

Interpattern association (IPA) model which the interconnection weight matrix(IWM) is constructed by the association between patterns is effective in similar pattern recognitions. But, if the number of reference patterns is increased, the ability of recognition is decreased. Using a hierarchical recognition algorithm which adopts the tree search strategy, we classified reference patterns into sub-groups by similarity. In IPA model, if input includes random noise we make it converge to reference pattern by means of input includes random noise we make it converge to reference pattern by means of increasing the number of pixels of prohibited state in IWM. In relation to reference patterns the pixel of prohibited state made partially prohibited state of no connected state using which is not included common and feature regions by each reference patterns.

• Journal of Communications and Networks
• /
• v.13 no.1
• /
• pp.63-69
• /
• 2011

In this paper, we propose a fixed-complexity sphere encoder (FSE) for multi-user multi-input multi-output (MU-MIMO) systems. The proposed FSE accomplishes a scalable tradeoff between performance and complexity. Also, because it has a parallel tree-search structure, the proposed encoder can be easily pipelined, leading to a tremendous reduction in the precoding latency. The complexity of the proposed encoder is also analyzed, and we propose two techniques that reduce it. Simulation and analytical results demonstrate that in a $4{\times}4$ MU-MIMO system, the proposed FSE requires only 11.5% of the computational complexity needed by the conventional QR decomposition with M-algorithm encoder (QRDM-E). Also, the encoding throughput of the proposed encoder is 7.5 times that of the QRDM-E with tolerable degradation in the BER performance, while achieving the optimum diversity order.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.12B
• /
• pp.1535-1541
• /
• 2009

In this paper, we suggest an efficient path searching algorithm that reduces the hop count when each node sends a data in ZigBee networks. As the hop count reduces, the network traffic is also reduces and leads to less energy consumption. This enables the sensor network live longer with limited node power. The proposed path searching algorithm consists of two sub-algorithms. One for upstream process and the other for downstream process. When a node selects its proper routing path, the node not only uses the information of the parent and child node, but it also uses the neighbor nodes for each node. In the simulation, we changed various network environment factors such as network parameters, number of nodes, and number of neighbor nodes and observed their performances. We compare the performance to the previous ZigBee Tree routing algorithm with separate two algorithms, the upstream and the downstream, and then compare the performance when all two algorithms are applied.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.48 no.2
• /
• pp.16-27
• /
• 2011

Embryo is a very early stage of the development of multicellular organism such as animals and plants. It is an important research target for studying ontogeny because the fundamental body system of multicellular organism is determined during an embryo state. Researchers in the developmental biology have a large volume of embryo image databases for studying embryos and they frequently search for an embryo image efficiently from those databases. Thus, it is crucial to organize databases for their efficient search. Hierarchical clustering methods have been widely used for database organization. However, most of previous algorithms tend to produce a highly skewed tree as a result of clustering because they do not simultaneously consider both the size of a cluster and the number of objects within the cluster. The skewed tree requires much time to be traversed in users' search process. In this paper, we propose a method that effectively organizes a large volume of embryo image data in a balanced tree structure. We first represent embryo image data as a similarity-based graph. Next, we identify clusters by performing a graph partitioning algorithm repeatedly. We check constantly the size of a cluster and the number of objects, and partition clusters whose size is too large or whose number of objects is too high, which prevents clusters from growing too large or having too many objects. We show the superiority of the proposed method by extensive experiments. Moreover, we implement the visualization tool to help users quickly and easily navigate the embryo image database.

• Journal of KIISE:Software and Applications
• /
• v.32 no.4
• /
• pp.247-257
• /
• 2005

Generally, the network topology design problem is characterized as a kind of NP-hard combinatorial optimization problem, which is difficult to solve with the classical method because it has exponentially increasing complexity with the augmented network size. In this paper, we propose the efficient approach with two phase that is comprised of evolutionary computation approach based on Prufer number(PN), which can efficiently represent the spanning tree, and a heuristic method considering 2-connectivity, to solve the highly reliable network topology design problem minimizing the construction cost subject to network reliability: firstly, to find the spanning tree, genetic algorithm that is the most widely known type of evolutionary computation approach, is used; secondly, a heuristic method is employed, in order to search the optimal network topology based on the spanning tree obtained in the first Phase, considering 2-connectivity. Lastly, the performance of our approach is provided from the results of numerical examples.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.44 no.1
• /
• pp.71-79
• /
• 2007

Recent advances in the internet infrastructure have resulted in a large number of huge Web sites and portals worldwide. These Web sites are being visited by various types of users in many different ways. Among all the web page access sequences from different users, some of them occur so frequently that may need an attention from those who are interested. We call them frequent access patterns and access sequences that can be frequent the candidate patterns. Since these candidate patterns play an important role in the incremental Web mining, it is important to efficiently generate, add, delete, and search for them. This thesis presents a novel tree structure that can efficiently store the candidate patterns and a related set of algorithms for generating the tree structure, adding new patterns, deleting unnecessary patterns, and searching for the needed ones. The proposed tree structure has a kind of the 3 dimensional link structure and its nodes are layered.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.6
• /
• pp.113-120
• /
• 2024

With the advancement of information and communication technology, we can easily generate various forms of data in our daily lives. To efficiently manage such a large amount of data, systematic classification into categories is essential. For effective search and navigation, data is organized into a tree-like hierarchical structure known as a category tree, which is commonly seen in news websites and Wikipedia. As a result, various techniques have been proposed to classify large volumes of documents into the terminal nodes of category trees. However, document classification methods using category trees face a problem: as the height of the tree increases, the number of terminal nodes multiplies exponentially, which increases the probability of misclassification and ultimately leads to a reduction in classification accuracy. Therefore, in this paper, we propose a new node expansion-based classification algorithm that satisfies the classification accuracy required by the application, while enabling detailed categorization. The proposed method uses a greedy approach to prioritize the expansion of nodes with high classification accuracy, thereby maximizing the overall classification accuracy of the category tree. Experimental results on real data show that the proposed technique provides improved performance over naive methods.