http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
히스토그램 처리방법에 의한 잡음 스펙트럼 추정을 이용한 잡음환경에서의 음성인식
권영욱,김형순,Kwon, Young-Uk,Kim, Hyung-Soon 한국음향학회 1997 韓國音響學會誌 Vol.16 No.5
스펙트럼 차감법은 잡음이 더해진 환경에서의 음성인시기에 널리 사용되는 전처리 방법이지만, 이를 위해서는 잡음의 스펙트럼을 잘 추정할 필요가 있다. 본 논문에서는 잡음 스펙트럼의 추정방법으로 히스토그램 처리방법을 사용한다. 이 방법은 음성/비음성 구간의 구분을 할 필요가 없으며 서서히 변화하는 잡음의 스펙트럼도 추정할 수 있다는 점에서 여타의 잡음 추정방법에 비해 장점을 지닌다. 다양한 SNR 조건하에서 유색 가우시안 잡음 및 실제 자동차 소음을 부가시킨 음성에 대해 화자독립 고립단어 인식 실험을 수행한 결과, 히스토그램 처리방법에 기반을 둔 스펙트럼 차감법의 인식성능이 초기 비음성구간의 스펙트럼 평균을 이용한 기존의 잡음 스펙트럼 추정방법에 비해 우수한 성능을 나타내었다. Spectral subtraction is widely-used preprocessing technique for speech recognition in additive noise environments, but it requires a good estimate of the noise power spectrum. In this paper, we employ the histogram technique for the estimation of noise spectrum. This technique has advantages over other noise estimation methods in that it does not requires speech/non-speech detection and can estimate slowly-varying noise spectra. According to the speaker-independent isolated word recognition in both colored Gaussian and car noise environments under various SNR conditions. Histogram-technique-based spectral subtraction method yields superier performance to the one with conventional noise estimation method using the spectral average of initial frames during non-speech period.
Le Fort I 골절단술을 통한 상악의 후상방 회전에 따른 상순과 비부의 연조직 변화
권영욱(Young-Wook Kwon),표성운(Sung-Woon Pyo),이 원(Won Lee),박재억(Je Uk Park) 대한구강악안면외과학회 2011 대한구강악안면외과학회지 Vol.37 No.6
Introduction: This study evaluate the soft tissue changes to the upper lip and nose after Le Fort I maxillary posterosuperior rotational movement. Materials and Methods: Twenty Skeletal class III patients, who had undergone bimaxillary surgery with a maxillary Le Fort I osteotomy and bilateral sagittal split ramus osteotomy, were included in the study. The surgical plan for maxilla was posterosuperior rotational movement, with the rotation center in the anterior nasal spine (ANS) of maxilla. Soft and hard tissue changes were measured by evaluating the lateral cephalograms obtained prior to surgery and at least 6 months after surgery. For cephalometric analysis, four hard tissue landmarks ANS, posterior nasal spine [PNS], A point, U1 tip), and five soft tissue landmarks (pronasale [Pn], subnasale [Sn], A Point, upper lip [UL], stomion superius [StmS]) were marked. A paired t test, Pearson s correlation analysis and linear regression analysis were used to evaluate the soft and hard tissue changes and assess the correlation. A P value <0.05 was considered significant. Results: The U1 tip moved 2.52±1.54 mm posteriorly in the horizontal plane (P<0.05). Among the soft tissue landmarks, Pn moved 0.97±1.1 mm downward (P<0.05), UL moved 1.98±1.58 mm posteriorly (P<0.05) and 1.18±1.85 mm inferiorly (P<0.05), and StmS moved 1.68±1.48 mm posteriorly (P<0.05) and 1.06±1.29 mm inferiorly (P<0.05). The ratios of horizontal soft tissue movement to the hard tissue were 1:0.47 for the A point and A point, and 1:0.74 for the U1 tip and UL. Vertically, the movement ratio between the A point and A point was 1:0.38, between U1 tip and UL was 1:0.83, and between U1 tip and StmS was 1:0.79. Conclusion: Posterosuperior rotational movement of the maxilla in Le Fort I osteotomy results in posterior and inferior movement of UL. In addition, nasolabial angle was increased. Nasal tip and base of the nose showed a tendency to move downward and showed significant horizontal movement. The soft tissue changes in the upper lip and nasal area are believed to be induced by posterior movement at the UL area.
신희진,권영욱,조성민,김창현,박재억,Shin, Hee-Jin,Kwon, Young-Wook,Cho, Sung-Min,Kim, Change-Hyen,Park, Je-Uk 대한악안면성형재건외과학회 2010 Maxillofacial Plastic Reconstructive Surgery Vol.32 No.5
Eagle's syndrome is the manifestation of elongated styloid process that is not a common entity in dental clinics. Without the knowledge of this syndrome misdiagnosis can be made and therefore differential diag-nosis with other diseases of oral and maxillofacial area is required. The symptoms of elongated styloid process are cervical pain, foreign body sensation, dysphagia and pharyngeal pain. A 27 years old male vis-ited our hospital. He had cervical pain with unilateral facial nerve palsy. Along with meticulous clinical examinations, 3D-CT, Plain radiographic studies were used to make the diagnosis. In computed tomogra-phy, Lt. styloid process was elongated and fractured. Patient have undergone surgical resection of elongat-ed styloid process and have been carefully observed. Characteristically, preoperative facial nerve palsy showed rapid postoperative recovery. The other clinical symptoms that were present in the initial visit were diminished. The purpose of this article is to report a case of Eagle's syndrome presenting unilateral facial nerve palsy that was treated with transoral surgical approach with literature review.
신태영,김기성,권영욱,김형순,Shin, Tae-Young,Kim, Gi-Sung,Kwon, Young-Uk,Kim, Hyung-Soon 한국음향학회 1997 韓國音響學會誌 Vol.16 No.6
Most of speech analysis methods developed up to date are based on second order statistics, and one of the biggest drawback of these methods is that they show dramatical performance degradation in noisy environments. On the contrary, the methods using higher order statistics(HOS), which has the property of suppressing Gaussian noise, enable robust feature extraction in noisy environments. In this paper we propose a text-independent speaker identification system using higher order statistics and compare its performance with that using the conventional second-order-statistics-based method in both white and colored noise environments. The proposed speaker identification system is based on the vector quantization approach, and employs HOS-based voiced/unvoiced detector in order to extract feature parameters for voiced speech only, which has non-Gaussian distribution and is known to contain most of speaker-specific characteristics. Experimental results using 50 speaker's database show that higher-order-statistics-based method gives a better identificaiton performance than the conventional second-order-statistics-based method in noisy environments. 음성 신호 처리에 널리사용되어 온 2차 통계에 의한 음성 분석 방법은 잡음 환경에서 성능이 크게 저하되는 단점을 지닌다. 이에 반하여 고차 통계 방법은 Gaussian 잡음 등을 억제하는 특성을 가지고 있어서 잡음 환경에 상대적으로 강인한 음성 특징 추출을 가능하게 한다. 본 논문에서는 고차 통계에 의한 음성 분석 방법을 이용하여 백색 및 유색 잡음 환경에서의 문맥 독립형(text-independent) 화자식별 시스템을 제안하고, 기존의 2차 통계에 의한 방식과 성능을 비교하였다. 본 논문에서의 화자식별 시스템은 벡터 양자화 방법에 기반을 두고 있으며, 고차 통계 방법에 의한 유성음/무성음 판별을 통해 non-Gaussian 특징을 가지면서도 화자 정보가 집중되어 있는 유성음 부분에 대해서만 음성 특징을 추출하여 인식에 사용하였다. 50명의 화자를 대상으로 한 화자식별 실험 결과, 고차 통계 방법이 2차 통계에 의한 방법보다 잡음 환경에서 상대적으로 우수한 인식 성능을 나타냄을 확인하였다.
전신질환자의 치과 치료 중/후 발생 가능한 합병증과 대처법
김창현(Change-Hyen Kim),신희진(Hee-Jin Shin),권영욱(Young-Wook Kwon),박재억(Je-Uk Park) 대한치과의사협회 2010 대한치과의사협회지 Vol.48 No.1
Nowadays, medically compromised patients who could not receive dental treatments in the past are able to go through minor oral surgeries with adequate preoperative measures. Thorough understanding of the systemic disease and its complications is needed as well as the management them. Frequent complications of surgical procedures are bleeding, infection, delayed healing, systemic reactions by stress and they can be aggravated due to the patients’ systemic conditions. Therefore, understanding of the systemic disease of patient visiting dental office and treatment modification according to the systemic status is needed. Also consultation to the medical doctor is imperative, through which perioperative risk and complications can be reduced. Among the high frequency complications of dental treatment of medically compromised patients, bleeding, infection, delayed healing, systemic reactions by stress will be discussed with the management of each one.
權榮郁,金一,金康彦,柳志久 釜山工業大學校 1994 論文集 Vol.36 No.-
One of the ultimate purposes in speech recognition is to recognize all of the arbitrary natural-sentences spoken by arbitrary speakers under any noisy environment. This paper describes how HMM is applied to speech recognition and specific parameters are estimated, and vector quantization(VQ) is obtained. Also some experiments of speech recognition are performed for digit that are pronounced four times by twenty Korean adults. The results show that recognition accuracy are 99.7% and 96.0% in closed-test recognition and open-test recognition, respectively.