• 한국지능시스템학회논문지
• /
• 제19권6호
• /
• pp.755-759
• /
• 2009

인간의 감정을 인식하는 기술은 인간-로봇 상호작용 분야의 중요한 연구주제 중 하나이다. 특히, 화자독립 감정인식은 음성감정인식의 상용화를 위해 꼭 필요한 중요한 이슈이다. 일반적으로, 화자독립 감정인식 시스템은 화자종속 시스템과 비교하여 감정특징 값들의 화자 그리고 성별에 따른 변화로 인하여 낮은 인식률을 보인다. 따라서 본 논문에서는 신뢰도 평가방법을 이용한 감정인식결과의 거절 방법을 사용하여 화자독립 감정인식 시스템을 일관되고 정확하게 구현할 수 있는 방법을 제시한다. 또한, 제안된 방법과 기존 방법의 비교를 통하여 제안된 방법의 효율성 및 가능성을 검증한다.

• 응용통계연구
• /
• 제16권2호
• /
• pp.377-393
• /
• 2003

본 논문에서는 두 모비율의 차에 대한 기존의 신뢰구간들을 소개하고 붓스트랩 신뢰구간도 제안하였다 또한 모비율의 차에 대한 신뢰구간이 가지는 성질로서 근사신뢰구간의 하향추정의 문제와 정확신뢰구간의 상향추정의 문제점들을 확인하였고 평균포함 확률, 구간기대폭 그리고 왜도성 측면에서 종합적인 비교를 하였다. 특히 모수에 대한 사전분포를 가정하여 여러 신뢰구간들이 지니는 특징도 살펴보았다 기존의 신뢰구간들과 제안된 붓스트랩 신뢰구간은 소표본의 모의실험을 통하여 실제 포함확률의 평균을 기준으로 비교되었고 이항분포에서와 같이 정확신뢰구간이 지니는 보수성을 확인할 수 있었다. 신뢰구간의 평균포함확률의 등고선 그림도 소개하였다.

• International Journal of Advanced Culture Technology
• /
• 제9권4호
• /
• pp.162-168
• /
• 2021

This is a group pretest-posttest design study that aim to evaluate the effect of problem-based learning using flipped learning to self leadership, nursing leadership, and confidence of nursing students. The subjects were 87 4th nursing students who took nursing management from March 15 to April 26, 2021. A questionnaire was used to measure self leadership, nursing leadership, and confidence. The measurement of self leadership is a 5-point Likert scale that employs a questionnaire. Flipped learning uploaded a 5-8 minute lecture video, and the subjects learned on learning management system, using smart device. PBL teaching was developed about personal growth, cooperation, nursing excellence, creative problem solving, influence. As a result of the study, self leadership was significantly improved after education than before education(t=-6.27, p<.001). Nursing leadership and confidence were significantly changed before and after education(t=-16.10, p<.001; t=-2.37, p<.001). According to the results of this study, the PBL teaching method using flipped learning is an effective teaching and learning strategy for improving leadership, and confidence. In future studies, it will be necessary to develope and verify specific teaching and learning methods by applying nursing experts in consideration of the type of leadership.

• 음성과학
• /
• 제9권1호
• /
• pp.49-59
• /
• 2002

This paper concerns an effective text-dependent speaker verification method to increase the performance of speaker verification. While various speaker verification methods have already been developed, their effectiveness has not yet been formally proven in terms of achieving acceptable performance levels. This paper proposes a weighted likelihood procedure along with a confidence measure based on subword-based text-dependent speaker verification. Our aim is to remedy the low performance problem in speaker verification by exploring a means to strengthen the verification likelihood via subword-based hypothesis criteria and weighted likelihood method. Experimental results show that the proposed speaker verification method outperforms that of the speaker verification scheme without using the proposed decision by a factor of up to 1.6 times. From these results, the proposed speaker verification method is shown to be very effective and to achieve a reliable performance.

• 대한음성학회지:말소리
• /
• 제46호
• /
• pp.103-115
• /
• 2003

In this paper, we proposed Background Model Set algorithm used in the speaker verification to improve calculating confidence measure(CM) in speech recognition. CM is to display relative likelihood between recognized models and antiphone models. In previous method calculating of CM, we calculated probability and standard deviation using all phonemes in composition of antiphone models. At this process, antiphone CM brought bad recognition result. Also, recognition time increases. In order to solve this problem, we studied about method to reconstitute average and standard deviation using BMS algorithm in CM calculation.

• 천문학회지
• /
• 제21권2호
• /
• pp.133-153
• /
• 1988

Linear polarization has been measured for fifteen extragalactic radio sources in the field of supernova remnant S 147. These observations were made at 4885, 4835, 1665, 1515, 1465, and 1385 MHz using the Very Large Array, primarily to determine unambiguous Rotation Measures (RM) of the sources. This yields a total of 11 new RM sources. Comparisons of a sample of sources which are seen through S 147 were made with sources located farther away. The result tentatively indicates that the distribution of the rotation measure of the former population is broadened, with more than a 85% level of confidence. This constitutes evidence that there is a SNR contribution to rotation measure in sources seen through the SNR. Limits on this RM contribution are statistically at 30 < | RM | < 70 rad $m^{-2}$ with a 85% statistical level of confidence. These result imply the magnetic field in the rim of S 147 has strength 10 microguass if the electron density is $1\;cm^{-3}$.

• 음성과학
• /
• 제15권4호
• /
• pp.85-96
• /
• 2008

Automatic speech recognition has severe performance degradation under noisy environments. To cope with the noise problem, many methods have been proposed. Most of them focused on noise-robust features or model adaptation. However, researchers have overlooked utterance verification (UV) under noisy environments. In this paper we discuss UV problems based on the normalized confidence measure. First, we show that UV performance is also degraded in noisy environments with the experiments of an isolated word recognition. Then we observe how the degradation of UV performances is suffered. Based on the UV experiments we propose a modeling method of the statistics of phone confidences using sigmoid functions. For obtaining the parameters of the sigmoidal models, the particle swarm optimization (PSO) is adopted. The proposed method improves 20% rejection performance. Our experimental results show that the PSO-NCM can apply noise speech recognition successfully.

• 품질경영학회지
• /
• 제18권1호
• /
• pp.40-47
• /
• 1990

The two-sample problem where an experimental treatment is compared with a control is considered. Without making any parametric-model assumptions for the distributions(or survival distributions), a measure for summarizing the differences between the treatment and the control is introduced. A method for constructing a confidence interval for the proposed measure is given in cases of complete and right random censored data. This method is illustrated with two numerical examples.

• 대한음성학회:학술대회논문집
• /
• 대한음성학회 2003년도 5월 학술대회지
• /
• pp.79-82
• /
• 2003

In this paper, we proposed Backgorund Model Set algorithm for the speaker verification to improve the shortcoming of calculating process in conventional confidence measure(CM). CM is to display relative likelihood between recognized models and unrecognized models. Unrecognized models is known as antiphone models. Calculate probability and standard deviation using all phonemes at process that compose antiphone model. At this process, antiphone CM brought bad result. Also, recognition time increases. In order problem, we studied about method to reconstitute average and standard deviation taking BMS algorithm using antiphoneme that near phoneme of CM calculation.

• 대한치위생과학회지
• /
• 제3권2호
• /
• pp.39-49
• /
• 2020

The purpose of this study was to provide the basic data needed to plan the effective teaching design of scaling education and to measure the changes in students' confidence in performing scaling according to the number of scaling practices. Scaling education is presented as a core basic hygiene competency of dental hygienists and evaluated as the practical competency of dental hygienists. This study selected 48 third-year students from the department of dental hygiene at S University in Asan, and analyzed data from completed participant surveys. The degree of "confidence in facing the subject in clinical practice," "confidence in applying hand instrument to the subject," and "confidence in applying ultrasonic instruments to the subject" was assessed on a five-point Likert scale. The difference in confidence was analyzed during five practice sessions. Each response was compared using frequency analysis, chi-square test, and repeated measurement ANOVA. Students who complained of a lack of confidence in the 14.6%, 25.0%, and 12.5%, respectively, in face-to-face practice, hand instruments, and ultrasonic instrument application responses. The more the practice was repeated, the more confident students were in all three categories, and the more statistically significant (p<0.001). As the number of scaling practice sessions increased, students' confidence in performing scaling also improved. In particular, the level of self-confidence was higher after the third practice session when compared to the first session. Therefore, it is necessary to design effective courses for teaching scaling practices so that at least three repetitive practice periods can be provided in clinical dental hygiene practices.