• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.30B no.2
• /
• pp.51-58
• /
• 1993

This paper describes the development of a speaker-dependent isolated word recognizer as applied to voice dialing in a car noise environment. for this purpose, several methods to improve performance under such condition are evaluated using database collected in a small car moving at 100km/h The main features of the recognizer are as follow: The endpoint detection error can be reduced by using the magnitude of the signal which is inverse filtered by the AR model of the background noise, and it can be compensated by using variants of the DTW algorithm. To remove the noise, an autocorrelation subtraction method is used with the constraint that residual energy obtainable by linear predictive analysis should be positive. By using the noise rubust distance measure, distortion of the feature vector is minimized. The speech recognizer is implemented using the Motorola DSP56001(24-bit general purpose digital signal processor). The recognition database is composed of 50 Korean names spoken by 3 male speakers. The recognition error rate of the system is reduced to 4.3% using a single reference pattern for each word and 1.5% using 2 reference patterns for each word.

• Proceedings of the Korea Committee for Ocean Resources and Engineering Conference
• /
• 2006.11a
• /
• pp.307-310
• /
• 2006

Underwater ambient noise originating from the geophysical, biological, and man-made acoustic sources contains much information on the sources and the ocean environment affecting the performance of the sonar equipments. In this paper, a set of feature vectors of the ambient noises using MFCC is proposed and extracted to form a data base for the purpose of identifying the noise sources. The developed algorithm for the pattern recognition is applied to the observed ocean data, and the initial results are presented and discussed.

• Journal of Biomedical Engineering Research
• /
• v.11 no.1
• /
• pp.157-162
• /
• 1990

In this study, a peripheral auditory model is used as a frequency feature extractor and a recurrent network which has recurrent links on input nodes is constructed in order to show the reliability of the recurrent network as a recognizer by executing recognition tests for 4 Korean place names and syllables. In the case of using the general learning rule, it is found that the weights are diverged for a long sequence because of the characteristics of the node function in the hidden and output layers. So, a refined weight compensation method is proposed and, using this method, it is possible to improve the system operation and to use long data. The recognition results are considerably good, even if time worping and endpoint detection are omitted and learning patterns and test patterns are made of average length of data. The recurrent network used in this study reflects well time information of temporal speech signal.

• Annual Conference of KIPS
• /
• 2018.10a
• /
• pp.272-275
• /
• 2018

금전적 이득을 극대화하기 위해 특정 기업, 기관을 대상으로 하는 공격이 증가하고 있다. 공격에 사용되는 일반적인 악성코드의 특징은 기존 시그니처 탐지 방법으로는 탐지하기 어렵다는 것과 공격자의 C&C 서버와의 통신이 일어난다는 점이다. 기업, 기관을 대상으로 이러한 악성코드를 이용한 공격이 시도된다면 감염된 모든 PC로부터 공통적인 네트워크 접속 기록이 발견될 수 있다. 따라서 본 논문에서는 이러한 특징에 중점을 두고 라이브 포렌식 오픈 소스를 활용하여 엔드포인트의 네트워크 접속 기록을 활용해 공통 행위를 탐지하는 기법을 제시하고자 한다.

• Genomics & Informatics
• /
• v.19 no.2
• /
• pp.13.1-13.8
• /
• 2021

Ranked in the topmost position among the deadliest diseases in the world, cardiovascular diseases (CVDs) are a global burden with alterations in heart and blood vessels. Early diagnostics and prognostics could be the best possible solution in CVD management. OMICS (genomics, proteomics, transcriptomics, and metabolomics) approaches could be able to tackle the challenges against CVDs. Genome-wide association studies along with next-generation sequencing with various computational biology tools could lead a new sight in early detection and possible therapeutics of CVDs. Human cardiac proteins are also characterized by mass spectrophotometry which could open the scope of proteomics approaches in CVD. Besides this, regulation of gene expression by transcriptomics approaches exhibits a new insight while metabolomics is the endpoint on the downstream of multi-omics approaches to confront CVDs from the early onset. Although a lot of challenges needed to overcome in CVD management, OMICS approaches are certainly a new prospect.

• Korean Journal of Veterinary Research
• /
• v.57 no.1
• /
• pp.31-36
• /
• 2017

Japanese encephalitis (JE) is an important zoonosis caused by the mosquito-transmitted JE virus (JEV), which is a causative agent of reproductive failure in pregnant sows. Detection of JEV antibodies in swine is performed by hemagglutination inhibition (HI), virus neutralization (VN), and the plaque reduction neutralization test (PRNT). The most stringent PRNT is the 90% endpoint PRNT ($PRNT_{90}$). These conventional assays are difficult to carry out in diagnostic laboratories with insufficient instruments or cell culture systems. An alternative assay that is easily conducted and time efficient is required. In this study, we improved the indirect enzyme-linked immunosorbent assay (I-ELISA) with clarified antigen for the detection of JEV antibodies. The I-ELISA results obtained from 175 swine serum samples were compared with HI, VN, and $PRNT_{90}$ results. The sensitivity of I-ELISA was 91.8%, 95.0%, and 94.7% compared with HI, VN, and $PRNT_{90}$ results, respectively. The specificity of I-ELISA was 92.2%, 94.7%, and 94.7% compared with HI, VN, and $PRNT_{90}$ results, respectively. Moreover, the I-ELISA results were significantly correlated with the HI (r = 0.93), VN (r = 0.95), and $PRNT_{90}$ (r = 0.92) results. These results suggest that the improved I-ELISA is useful for serosurveillance of JEV in swine.

• Journal of Gastric Cancer
• /
• v.19 no.2
• /
• pp.157-164
• /
• 2019

Purpose: Although standard radical gastrectomy is recommended after noncurative resection of endoscopic submucosal dissection (ESD) for early gastric cancer in most cases, residual tumor and lymph node metastasis have not been identified after surgery. The aim of this study is to evaluate the feasibility of sentinel node navigation surgery after noncurative ESD. Materials and Methods: This trial is an investigator-initiated, multicenter prospective phase II trial. Patients who underwent ESD for clinical stage T1N0M0 gastric cancer with noncurative resections were eligible. Qualified investigators who completed the prior phase III trial (SENORITA 1) are exclusively allowed to participate. In this study, 2 detection methods will be used: 1) intraoperative endoscopic submucosal injection of dual tracer, including radioisotope and indocyanine green (ICG) with sentinel basins detected using gamma-probe; 2) endoscopic injection of ICG, with sentinel basins detected using a fluorescence imaging system. Standard laparoscopic gastrectomy with lymphadenectomy will be performed. Sample size is calculated based on the inferior confidence interval of the detection rate of 95%, and the calculated accrual is 237 patients. The primary endpoint is detection rate, and the secondary endpoints are sensitivity and postoperative complications. Conclusions: This study is expected to clarify the feasibility of laparoscopic sentinel basin dissection after noncurative ESD. If the feasibility is demonstrated, a multicenter phase III trial will be initiated to compare laparoscopic sentinel node navigation surgery versus laparoscopic standard gastrectomy in early gastric cancer after endoscopic resection.

• Clinical Endoscopy
• /
• v.56 no.3
• /
• pp.315-324
• /
• 2023

Background/Aims: Image-enhanced endoscopy can detect superficial oro-hypopharyngeal squamous cell carcinoma; however, reliable endoscopy of the pharyngeal region is challenging. Endoscopy under general anesthesia during transoral surgery occasionally reveals multiple synchronous lesions that remained undetected on preoperative endoscopy. Therefore, we aimed to determine the lesion detection capability of endoscopy under general anesthesia for superficial oro-hypopharyngeal squamous cell carcinoma. Methods: This retrospective study included 63 patients who underwent transoral surgery for superficial oropharyngeal squamous cell carcinoma between April 2005 and December 2020. The primary endpoint was to compare the lesion detection capabilities of preoperative endoscopy and endoscopy under general anesthesia. Other endpoints included the comparison of clinicopathological findings between lesions detected using preoperative endoscopy and those newly detected using endoscopy under general anesthesia. Results: Fifty-eight patients (85 lesions) were analyzed. The mean number of lesions per patient detected was 1.17 for preoperative endoscopy and 1.47 for endoscopy under general anesthesia. Endoscopy under general anesthesia helped detect more lesions than preoperative endoscopy did (p<0.001). The lesions that were newly detected on endoscopy under general anesthesia were small and characterized by few changes in color and surface ruggedness. Conclusions: Endoscopy under general anesthesia for superficial squamous cell carcinoma is helpful for detecting multiple synchronous lesions.

• Biomolecules & Therapeutics
• /
• v.13 no.1
• /
• pp.59-63
• /
• 2005

A kinetic assay was carried out in order to compare the ability of detection for prekallikrein activator(PKA) in plasma-derived products with that of an endpoint assay and a commercial method. Using these methods, 9 human albumin preparations were assayed and compared to each other. The coefficient of variation between the Kinetic assay and the end point assay was found within 6.6% and this result showed that two methods were highly correlative and the end point assay could act as a replacement of the kinetic assay. Another important goal of this study was to investigate the reproducibility among laboratories on the kinetic assay. A collaborative study was performed to validate the kinetic method with intra and inter assays. The coefficient of variation for the intra assay of each laboratory was less than 4% and that for between individuals in the inter assay was 4.1%. These results revealed that the kinetic assay showed good reproducibility. The contents of PKA in plasma-derived products were also determined by the kinetic assay. As a result, it was found that trace amounts of PKA were present in 32 human immunoglobulin preparations, however the average concentration of PKA in 171 albumin preparations was 5.8 IU/mL.

• The Journal of the Acoustical Society of Korea
• /
• v.17 no.4
• /
• pp.14-22
• /
• 1998

본 연구에서는 Texas Instruments 사의 저가형 부동소수점 디지털 신호 처리기 (Digital Singnal Processor, DSP)인 TMS320C32를 이용하여 실시간 화자종속 음성인식 하 드웨어 모듈(VR32)을 개발하였다. 하드웨어 모듈의 구성은 40MHz의 TMS320C32 DSP, 14bit 코덱인 TLC32044(또는 8bit μ-law PCM 코덱), EPROM과 SRAM 등의 메모리와 호 스트 인터페이스를 위한 로직 회로로 이루어졌다. 뿐만 아니라 이 하드웨어 모듈을 PC사에 서 평가해보기 위한 PC 인터페이스용 보드 및 소프트웨어도 개발하였다. 음성인식 알고리 즘의 구성은 에너지와 ZCR을 기반으로 한 끝점검출(Endpoint Detection) 침 10차 가중 LPC 켑스터럼(Weighted LPC Cepstrum) 분석이 실시간으로 이루어지며 이후 Dynamic Time Warping(DTW)를 통하여 최고 유사 단어를 결정하고 다시 검증과정을 거쳐 최종 인식을 수행한다. 끝점검출의 경우 적응 문턱값(Adaptive threshold)을 이용하여 잡음에 강인한 끝 점검출이 가능하며 DTW 알고리즘의 경우 C 및 어셈블리를 이용한 최적화를 통하여 계산 속도를 대폭 개선하였다. 현재 인식률은 일반 사무실 환경에서 통상 단축다이얼 용도로 사 용할 수 있는 30 단어에 대하여 95% 이상으로 매우 높은 편이며, 특히 배경음악이나 자동 차 소음과 같은 잡음환경에서도 잘 동작한다.