• 대한음성학회지:말소리
• /
• 제58호
• /
• pp.67-81
• /
• 2006

For a large corpus of time-aligned data, HMM based approaches are most widely used for automatic segmentation, providing a consistent and accurate phone labeling scheme. There are two methods for training in HMM. Flat starting method has a property that human interference is minimized but it has low accuracy. Bootstrap method has a high accuracy, but it has a defect that manual segmentation is required In this paper, a new algorithm is proposed to minimize manual work and to improve the performance of automatic segmentation. At first phase, voiced, unvoiced and silence classification is performed for each speech data frame. At second phase, the phoneme sequence is aligned dynamically to the voiced/unvoiced/silence sequence according to the acoustic phonetic rules. Finally, using these segmented speech data as a bootstrap, phoneme model parameters based on HMM are trained. For the performance test, hand labeled ETRI speech DB was used. The experiment results showed that our algorithm achieved 10% improvement of segmentation accuracy within 20 ms tolerable error range. Especially for the unvoiced consonants, it showed 30% improvement.

• 임상간호연구
• /
• 제25권2호
• /
• pp.198-207
• /
• 2019

Purpose: The purpose of this study was to provide data to establish an efficient manpower management plan so that the organization can identify the influence of organizational silence and professionalism on job embeddedness among nurses in hospital management. Methods: Self-report questionnaires were distributed to 150 nurses in hospital management who were working in general hospitals with less than 900 beds but more than 300 beds in the districts of Seoul and Gyunggi, Gangwon province. Of the questionnaires, 127 out of 150 copies were collected and 120 copies were used for final analysis. Seven were not considered sincere in the responses. Data were analyzed using SPSS ver. 22.0 statistic program, t-test, ANOVA, $Scheff{\acute{e}}$ test, Pearson correlation coefficients and multiple regression. Results: Organizational silence and professionalism had significant influence on job embeddedness, which explained 43.4% of job embeddedness. Conclusion: The results indicate that in order to raise job embeddedness, it is necessary to provide various programs to change communication patterns in the nurses in hospital management, and to create environments in which nurses in hospital management can present various opinions.

• Safety and Health at Work
• /
• 제12권2호
• /
• pp.192-200
• /
• 2021

Background: A number of accidents have occurred in small and medium enterprises (SMEs). Efforts in reducing accidents have been undertaken through the implementation of safety behaviors. Unfortunately, few studies have examined motives behind unsafe behaviors, such as safety silence motives. This study aimed to observe the motives underlying safety behaviors, namely safety silence motive (SSM) (SSM-relation, SSM-climate, SSM-issue, and SSM-job) and to evaluate the effect of SSM and safety communication on safety participation in different industrial sectors and scales. Materials and Methods: Eighty workers from two industrial sectors and scales of SMEs were involved. They were instructed to fill out a set of questionnaires. A five-Likert scale was used to respond. An independent t test was applied to find any significant differences. The partial least square-structural equation modeling for multigroup was used to develop a model on relations among the variables. Results: The results showed that SSM scores were high in SMEs, and the scores were different across industrial sectors and scales. SSM had a negative influence on safety communication, and safety communication positively influenced safety participation. Conclusion: The study of SSM, safety communication, and safety participation in different sectors and scales should be separated in SMEs.

• 디지털융복합연구
• /
• 제16권3호
• /
• pp.385-394
• /
• 2018

본 연구의 목적은 종합병원에 근무하는 간호사의 병원조직문화, 조직침묵과 직무 배태성이 이직의도에 미치는 영향을 확인하기 위함이다. 연구대상자는 종합병원 근무 간호사 152명이었으며 자료수집은 2017년 4월 19일부터 4월 30일까지 수집하였다. 통계분석은 SPSS/WIN 22.0 프로그램을 이용하여 위계적 다중 회귀분석으로 분석하였다. 연구결과, 이직의도는 체념적, 방어적 침묵이 정적 상관관계를 보인 반면, 합의지향적 조직문화, 직무배태성의 희생, 적합성, 연계가 부적 상관관계를 나타냈다. 이직의도의 영향요인으로는 현 근무지 경력, 방어적 침묵, 희생이 있었으며, 이직의도에 대한 이들의 설명력은 47%로 나타났다. 따라서 이직의도를 낮추기 위해서 적절한 부서이동의 기회 및 간호조직 내 커뮤니케이션 패턴 변화를 위한 다양한 프로그램 제공, 타 기관과 차별화된 복지 제도 마련이 필요할 것으로 사료된다.

• 한국융합학회논문지
• /
• 제8권5호
• /
• pp.13-17
• /
• 2017

음성 인식에서 기존의 음성 특징 추출 방법은 명확하지 않은 스레숄드 값으로 인해 부정확한 음성 인식률을 가진다. 본 연구에서는 음성과 비음성에 대한 특징 추출을 묵음 특징 정규화를 융합한 음성 인식 성능 향상을 위한 방법을 모델링 한다. 제안한 방법에서는 잡음의 영향을 최소화하여 모델을 구성하였고, 각 음성 프레임에 대해 음성 신호 특징을 추출하여 음성 인식 모델을 구성하였고, 이를 묵음 특징 정규화를 융합하여 에너지 스펙트럼을 엔트로피와 유사하게 표현하여 원래의 음성 신호를 생성하고 음성의 특징이 잡음을 적게 받도록 하였다. 셉스트럼에서 음성과 비음성 분류의 기준 값을 정하여 신호 대 잡음 비율이 낮은 신호에서 묵음 특징 정규화로 성능을 향상하였다. 논문에서 제시하는 방법의 성능 분석은 HMM과 CHMM을 비교하여 결과를 보였으며, 기존의 HMM과 CHMM을 비교한 결과 음성 종속 단계에서는 2.1%p의 인식률 향상이 있었으며, 음성 독립 단계에서는 0.7%p 만큼의 인식률 향상이 있었다.

• 한국통신학회논문지
• /
• 제35권7C호
• /
• pp.604-610
• /
• 2010

본 논문에서는 기존의 CLSFN (Cepstral distance and Log-energy based Silence Feature Normalization) 방법의 인식성능을 향상시키기 위하여, 필터 뱅크 서브 밴드 영역에서 잡음을 차감하는 방법과 CLSFN을 결합하는 방법, 즉 FSFN (Filter bank sub-band energy subtraction based CLSFN)을 제안하였다. 이 방법은 음성으로부터 특징 파라미터를 추출할 때 필터 뱅크 서브 밴드 영역에서 잡음을 제거하여 켑스트럼 특징을 향상시키고, 이에 대한 켑스트럼 거리를 이용하여 음성/묵음 분류의 정확도를 개선함으로써 기존 CLSFN 방법에 비해 향상된 인식성능을 얻을 수 있다. Aurora 2.0 DB를 이용한 실험결과, 제안하는 FSFN 방법은 CLSFN 방법에 비해 평균 단어 정확도 (word accuracy)가 약 2% 향상되었으며, CMVN (Cepstral Mean and Variance Normalization)과의 결합에서도 기존 모든 방법에 비해 가장 우수한 인식성능을 나타내어 제안 방법의 유효성을 확인할 수 있었다.

• 한국산업정보학회논문지
• /
• 제22권4호
• /
• pp.131-147
• /
• 2017

본 연구는 진성리더십이 상사에 대한 신뢰와 조직침묵을 매개로 이직의도와 반생산적 과업행동에 미치는 영향을 통합적으로 분석하고자 하며 이를 통하여 진성리더십의 중요성과 필요성을 강조하고자 한다. 부산 소재 중규모 제조업체 종업원들을 대상으로 설문조사를 실시하였다. 실증분석 결과 첫째, 진성리더십의 구성요소를 상사의 투명성, 도덕성, 균형 잡힌 처리, 자아인식으로 구분되었으며 구성요소 모두 상사에 대한 신뢰에 정(+)의 유의미한 영향을 미치는 것으로 나타났다. 둘째, 상사에 대한 신뢰는 구성원들의 조직침묵(체념적, 방어적)에 부(-)의 유의미한 영향을 미치는 것으로 나타났으며 셋째, 조직침묵은 조직 내 구성원들의 이직의도와 반생산적 과업행동에 모두 정(+)의 유의미한 영향을 미치는 것으로 나타났다.

• Molecules and Cells
• /
• 제42권5호
• /
• pp.397-405
• /
• 2019

The regulatory role of long noncoding RNA (lncRNA) growth arrest-specific transcript 5 (GAS5) in both cancerous and noncancerous cells have been widely reported. This study aimed to evaluate the role of lncRNA GAS5 in heart failure caused by myocardial infarction. We reported that silence of lncRNA GAS5 attenuated hypoxia-triggered cell death, as cell viability was increased and apoptosis rate was decreased. This phenomenon was coupled with the down-regulated expression of p53, Bax and cleaved caspase-3, as well as the up-regulated expression of CyclinD1, CDK4 and Bcl-2. At the meantime, the expression of four heart failure-related miR-NAs was altered when lncRNA GAS5 was silenced (miR-21 and miR-142-5p were up-regulated; miR-30b and miR-93 were down-regulated). RNA immunoprecipitation assay results showed that lncRNA GAS5 worked as a molecular sponge for miR-142-5p. More interestingly, the protective actions of lncRNA GAS5 silence on hypoxia-stimulated cells were attenuated by miR-142-5p suppression. Besides, TP53INP1 was a target gene for miR-142-5p. Silence of lncRNA GAS5 promoted the activation of PI3K/AKT and MEK/ERK signaling pathways in a miR-142-5p-dependent manner. Collectively, this study demonstrated that silence of lncRNA GAS5 protected H9c2 cells against hypoxia-induced injury possibly via sponging miR-142-5p, functionally releasing TP53INP1 mRNA transcripts that are normally targeted by miR-142-5p.

• 대한음성학회:학술대회논문집
• /
• 대한음성학회 2005년도 춘계 학술대회 발표논문집
• /
• pp.49-52
• /
• 2005

Cepstral Mean Subtraction (CMS) makes effectively compensation for a channel distortion, but there are some shortcomings such as distortions of feature parameters, waiting for the whole speech sentence. By assuming that the silence parts have the channel characteristics, we consider the channel normalization using subtraction of cepstral means which are only obtained in the silence areas. If the considered techniques are successfully used for the channel compensation, the proposed method can be used for real time processing environments or time important areas. In the experiment result, however, the performance of our method is not good as CMS technique. From the analysis of the results, we found potentiality of the proposed method and will try to find the technique reducing the gap between CMS and ours method.

• Journal of Magnetics
• /
• 제22권1호
• /
• pp.85-92
• /
• 2017

This research is proposed with the efficient searching method of deperming protocol for magnetic silence. The deperming protocol has been changed depending on the initial magnetic field, the final magnetic field, and the shot number, so deciding parameter is highly important. Therefore, in this paper, the value of the initial magnetic field is supposed to be fixed to that of the saturation field, and excluding the initial magnetic field in the variable, the deperming protocol has been analyzed depending on the final magnetic field and the change of the shot number. In the result of this experiment, the final magnetic field gets high performance to the reverse of the direction of the initial magnetization, and it is resulted that the shot number inverse proportion to the size of the final magnetic field.