• Journal of Information Processing Systems
• /
• 제16권2호
• /
• pp.343-359
• /
• 2020

In this paper, a new algorithm is proposed for three-dimensional (3D) shape recognition using local features of model views and its sparse representation. The algorithm starts with the normalization of 3D models and the extraction of 2D views from uniformly distributed viewpoints. Consequently, the 2D views are stacked over each other to from view cubes. The algorithm employs the descriptors of 3D local features in the view cubes after applying Gabor filters in various directions as the initial features for 3D shape recognition. In the training stage, we store some 3D local features to build the prototype dictionary of local features. To extract an intermediate feature vector, we measure the similarity between the local descriptors of a shape model and the local features of the prototype dictionary. We represent the intermediate feature vectors of 3D models in the sparse domain to obtain the final descriptors of the models. Finally, support vector machine classifiers are used to recognize the 3D models. Experimental results using the Princeton Shape Benchmark database showed the average recognition rate of 89.7% using 20 views. We compared the proposed approach with state-of-the-art approaches and the results showed the effectiveness of the proposed algorithm.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권7호
• /
• pp.2785-2799
• /
• 2020

We propose a novel speaker adaptation method using acoustic model clustering. The similarity of different speakers is defined by the cosine distance between their i-vectors (intermediate vectors), and various efficient clustering algorithms are applied to obtain a number of speaker subsets with different characteristics. The speaker-independent model is then retrained with the training data of the individual speaker subsets grouped by the clustering results, and an unknown speech is recognized by the retrained model of the closest cluster. The proposed method is applied to a large-scale speech recognition system implemented by a hybrid hidden Markov model and deep neural network framework. An experiment was conducted to evaluate the word error rates using Resource Management database. When the proposed speaker adaptation method using i-vector based clustering was applied, the performance, as compared to that of the conventional speaker-independent speech recognition model, was improved relatively by as much as 12.2% for the conventional fully neural network, and by as much as 10.5% for the bidirectional long short-term memory.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권12호
• /
• pp.5643-5656
• /
• 2016

The detection accuracy of steganalysis depends on many factors, including the embedding algorithm, the payload size, the steganalysis feature space and the properties of the cover source. In practice, the cover source mismatch (CSM) problem has been recognized as the single most important factor negatively affecting the performance. To address this problem, we propose a new framework for blind, universal steganalysis which uses traditional steganalyst features. Firstly, cover images with the same statistical properties are searched from a reference image database as aided samples. The test image and its aided samples form a whole test set. Then, by assuming that most of the aided samples are innocent, we conduct outlier detection on the test set to judge the test image as cover or stego. In this way, the framework has removed the need for training. Hence, it does not suffer from cover source mismatch. Because it performs anomaly detection rather than classification, this method is totally unsupervised. The results in our study show that this framework works superior than one-class support vector machine and the outlier detector without considering the image retrieval process.

• 한국지반공학회:학술대회논문집
• /
• 한국지반공학회 2009년도 춘계 학술발표회
• /
• pp.456-463
• /
• 2009

The preconsolidation pressure has been commonly determined by oedometer test. However, it can also be determined by in-situ test, such as piezocone test with theoretical and(or) empirical correlations. Recently, Neural Network(NN) theory was applied and some models were proposed to estimate the preconsolidation pressure or OCR. However, since the optimization process of synaptic weights of NN model is dependent on the initial synaptic weights, NN models which are trained with different initial weights can't avoid the variability on prediction result for new database even though they have same structure and use same transfer function. In this study, Committee Neural Network(CNN) model is proposed to improve the initial weight dependency of multi-layered neural network model on the prediction of preconsolidation pressure of soft clay from piezocone test result. It was found that even though the NN model has the optimized structure for given training data set, it still has the initial weight dependency, while the proposed CNN model can improve the initial weight dependency of the NN model and provide a consistent and precise inference result than existing NN models.

• 한국주거학회논문집
• /
• 제28권2호
• /
• pp.35-45
• /
• 2017

This study aimed to explore housing welfare service situation in City of Cheongju, Chungbuk, and assess demands for the service improvement through interviews of public and private field specialists in order to seek practical implications for enhancement of housing welfare policy and system in Cheongju. In December 2016, a series of personal and focus group interviews using semi-structured questionnaires were conducted to total 10 specialists who were currently providing and housing services. Major findings and implications are as follow. (1) City's own housing survey and database construction are in need for development and implementation of housing welfare service programs serving local needs effectively and efficiently. (2) The field specialists perceived importance of realignment of service system and to establishment of the Cheongju housing welfare center as a control tower of the services. (3) It is required to provide various opportunities to strengthen the specialists' competence through education, training and networking and to arrange persons in charge of housing programs in every community-level self-government. (4) The specialists showed their strong interests in introduction of diverse housing welfare services and programs.

• 한국항만경제학회지
• /
• 제17권1호
• /
• pp.73-90
• /
• 2001

Logistics has become a key element in a shipping firm's overall strategy, now that the concept of logistics has expanded into one that is more comprehensive and integrated. furthermore, the global shipping market is the most competitive market in the world. In this environment, shipping firms are using Logistics Information Systems (LIS), which help them to save time and reduce costs. As such, LIS has become a critical part of Korea's attempts to develop Its national logistics system and industry. In this light, this study attempted to empirically determine the critical elements necessary for effective shipping LIS implementation and to then analyze how these elements are related to actual logistics performance within the shipping industry. The research results show that the sample firms considered the following factors to be most Important to an effective LIS: the support of top management; an efficient and well-structured communication channel; a solid strategic implementation plan; and a high level of information analysis internally. Other factors mentioned as being somewhat significant included MIS organization and human resources, MIS discipline and training, software development capabilities, MIS organizational alignment, database management systems information standardization, competitive advantage, financial support and government support The main problems with current shipping LIS include the expensive cost of systems, lack of standardized working processes, and weak organizational structures.

• 한국음향학회지
• /
• 제19권5호
• /
• pp.9-12
• /
• 2000

음성 다이얼링 시스템은 화자의 음성을 인식하여 원하는 전화번호로 자동으로 전화를 걸어주는 시스템으로 주로 이동 전화나 휴대형 통신 장비에 유용하게 사용된다. 개인 음성 다이얼링 시스템의 경우, 다이얼링에 사용되는 모든 구문은 사용자가 선택하고 사용자의 음성을 사용하여 학습되어 음성 인식을 위한 HMM을 생성한다. 이러한 시스템은 화자독립 시스템 보다 매우 적은 메모리 공간과 계산량으로 구현이 가능하다. 그러나 이러한 시스템은 학습시 각 단어당 2-3개의 음성만을 사용하므로 음성인식 시스템의 성능을 개선하기 위한 각 상태에서의 상태지속분포을 추정하기는 매우 어렵다. 따라서 본 논문에서는 성능개선을 위한 후처리기를 제안하였다. 전화선을 통하여 구성된 데이터베이스를 이용한 실험에서 제안된 후처리기가 인식 시스템의 성능을 향상시킴을 확인하였다.

• Bulletin of the Korean Chemical Society
• /
• 제28권4호
• /
• pp.561-566
• /
• 2007

Adenosine kinase (AK) is a ubiquitous intracellular enzyme, which catalyzes the phosphorylation of adenosine (ADO) to adenosine monophosphate (AMP). AK inhibitors have therapeutic potential as analgesic and antiinflammatory agents. A chemical feature based pharmacophore model has been generated from known AK inhibitors (26 training set compounds) by HypoGen module implemented in CATALYST software. The top ranked hypothesis (Hypo1) contained four features of two hydrogen-bond acceptors (HBA) and two hydrophobic aromatics (Z). Hypo1 was validated by 124 test set molecules with a correlation coefficient of 0.905 between experimental and estimated activity. It was also validated by CatScramble method. Thus, the Hypo1 was exploited for searching new lead compounds over 238,819 chemical compounds in NCI database and then the selected compounds were screened based on restriction estimated activity and Lipinski's rules to evaluate their drug-like properties. Finally we could obtain 72 new lead candidates and the two best compound structures from them were posted.

• 대한한방부인과학회지
• /
• 제33권4호
• /
• pp.1-22
• /
• 2020

Objectives: The purpose of this review is to evaluate the efficacy of moxibustion for stress urinary incontinence (SUI) in women. Methods: For relevant randomized controlled trials (RCTs), we searched the following databases from their inception to September 1, 2020: The Cochrane Library, PubMed, EMBASE, Chinese National Knowledge Infrastructure Database (CNKI), Koreanstudies Information Service System (KISS), Research Information Sharing Service (RISS), and National Digital Science Library (NDSL). The key search terms were 'stress urinary incontinence' and 'moxibustion'. Data extraction and assessment of risk of bias were conducted by two authors independently. Results: A total of 11 RCTs were finally included in this systematic review. In all studies, moxibustion treatment was applied as an adjuvant therapy to the conventional treatment, and the most common conventional treatment was pelvic floor muscle training (PFMT). The treatment group (conventional treatment plus moxibustion) showed statistically more significant effect than the control group (conventional treatment only) in various evaluation indicators including urinary incontinence frequency, 1 hour urine pad test, quality of life, the clinical efficacy rate, and pelvic muscle strength. Conclusions: In this study, we investigated the efficacy of moxibustion as an adjuvant therapy for female SUI patients. Further studies are needed to supplement the safety of moxibustion and the evaluation of moxibustion dose.

• 한국콘텐츠학회논문지
• /
• 제4권3호
• /
• pp.107-113
• /
• 2004

본 논문은 얼굴의 기하학적인 특징과 웨이브릿 변환을 사용한 PCA/LDA 복합 방법을 제안하여 얼굴 인식 시스템의 성능을 향상시켰다. 기존의 PCA/LDA 방법은 형태적인 분산의 정도에 따라 유사도를 측정하였기 때문에 얼굴 윤곽선을 정확하게 반영하지 못하였다. 이 단점을 극복하기 위하여 본 논문에서는 눈과 입사이의 거리를 측정하여 질의영상과 훈련영상에서 큰 차이가 있을 경우에는 얼굴내의 눈, 코, 턱 각각의 영역에 대한 에너지를 특징 벡터로 사용하여 기즌의 PCA/LDA로 계산한 유사도를 재산정하였다. 본 논문에서 제안한 방법을 이용해서 ORL 데이터베이스의 400개 얼굴 영상에 대해 모의 실험한 결과 기존의 PCA/LDA 방법보다 약 4％의 인식률 향상이 있음을 보였다