• ETRI Journal
• /
• 제38권2호
• /
• pp.347-355
• /
• 2016

In face examinations, gender classification (GC) is one of several fundamental tasks. Recent literature on GC primarily utilizes datasets containing high-resolution images of faces captured in uncontrolled real-world settings. In contrast, there have been few efforts that focus on utilizing low-resolution images of faces in GC. We propose a GC method based on a pixel classifier boosting with modified census transform features. Experiments are conducted using large datasets, such as Labeled Faces in the Wild and The Images of Groups, and standard protocols of GC communities. Experimental results show that, despite using low-resolution facial images that have a 15-pixel inter-ocular distance, the proposed method records a higher classification rate compared to current state-of-the-art GC algorithms.

• 한국음향학회지
• /
• 제27권5호
• /
• pp.252-255
• /
• 2008

본 논문에서는 성별 인식 시스템의 성능향상을 위해 변별적 가중치 학습 (discriminative weight training) 기반의 최적화된 SVM (support vector machine)을 제안한다. MCE (minimum classification error)방법을 도입하여, 각각의 MFCC (mel-frequency cepstral coefficients) 특징벡터 차수별로 다른 가중치를 가지는 SVM을 제안한다. 제안된 알고리즘은 기존의 동일 가중치를 가지는 SVM 기반의 성별인식 시스템과 비교하였으며, 우수한 성능을 보인다.

• 융합신호처리학회논문지
• /
• 제12권2호
• /
• pp.113-118
• /
• 2011

본 논문에서는 무선센서네트워크에서 이루어지는 협동적 센서융합을 이용한 화자성별분류를 제안하였다. 센서노드들은 BER(Band Energy Ratio) 기반 음성활동검출을 수행함으로써 불필요한 입력 데이터는 제거하고 관련성이 높은 데이터만을 처리 및 경판정한다. 개별적 센서노드에서 생성된 경판정 값들은 융합센터로 송신되고 전역적 결정 융합을 구축하기 때문에 전력 소모를 줄이고 네크워크 자원을 절약한다. 화자성별분류를 위한 센서융합기법으로써 베이시안(Bayesian) 센서융합 및 전역적 가중결정융합가법들이 제안되었다. 베이시안 센서융합의 경우, 배치되는 센서노드 수 변화에 따른 ROC(Receiver Operating Characteristic) 커브의 동작점을 통해 개별 센서노드 레벨에서 얻어진 경판정 값들을 처리하고 최적의 분류 융합을 결정한다. 전역적 결정을 위한 가중치로써 BER 및 MCL(Mutual Confidence Level)을 채택하여 개별적 지역 경판정 값들을 효율적으로 결합 및 융합시킨다. 센서 노드의 수가 증가함에 따라 분류화 성능이 개선되어졌으며 특히 낮은 SNH(Signal to Noise Ratio) 환경에서 성능 개선폭이 더 높게 나타남을 실험적으로 확인하였다.

• 대한기계학회논문집A
• /
• 제41권11호
• /
• pp.1035-1045
• /
• 2017

얼굴 모양 및 목소리를 이용하는 방법을 포함하여 연령 및 성별을 분류하는 다양한 방법이 연구되고 있다. 그러나 얼굴 기반 방법은 원거리에서 인식률이 급격히 감소하고, 오디오 기반 방법은 잡음이 많은 환경에서는 적용하기 어렵다. 대조적으로 보행 기반 방법은 대상자가 카메라에 촬영만 되면 인식이 가능하다. 기존 연구에서 카메라의 시점은 측면에서만 볼 수 있어서 실제 환경에서 일반 보행과는 현실적으로 차이가 발생했다. 본 연구에서는 일반 보행 데이터를 이용하여 연령과 성별을 분류할 수 있도록 RGB-D 센서로부터 획득된 골격 모델을 이용한 특징 추출 방법을 제안한다. 실험 결과는 제안된 방법이 실제 환경에서 효율적임을 보여준다.

• 말소리와 음성과학
• /
• 제1권2호
• /
• pp.163-171
• /
• 2009

This study attempted to investigate vocal characteristics and differences in gender and voice classification among classical singers. Twenty-three female singers (M = 23.1 yrs, SD = 3.6 yrs, average 6.3 yrs singing experience, all classified as sopranos) and twenty male singers (M = 25.2 yrs, SD= 3.6 yrs, average 6. 3 yrs singing experience, 8 tenors, 12 baritones) were recruited to participate in the present study. Speaking fundamental frequency (FO), closed quotient (CQ), MPT (Maximum Phonation Time), breathing types, maximum inspiratory pressure (MIP), maximum expiratory pressure (MEP), and singers' formants were measured. In addition, vibratory patterns were observed using stroboscopy. Sfo, singing CQ, breathing types, formant frequency in singers' formants, MIP, MEP, and MPT were significantly different from gender to gender. Generally, singers' formants were observed in male singers and also the pattern of singers' formants was different between tenors and baritones. Lower singing CQ values were observed than speaking CQ values in the female singers (P<.001). Furthermore, MEP, MIP, and singing CQ were significantly lower for female singers than for males singers (P<.001). MPT and speaking FO, however, were not significantly different between tenors and baritones.

• 한국농업기계학회:학술대회논문집
• /
• 한국농업기계학회 2017년도 춘계공동학술대회
• /
• pp.23-23
• /
• 2017

Recent favor on high quality food and concern on food safety have demonstrated the superiority of Hanwoo (Korean native cattle). In general, the price of cow is higher than those of steer and bull, causing cheating issues in the market. Hence, this study is to discriminate genders of Hanwoo with identification of factors which highly influence gender discrimination based on the big-size deboning data. Totally, there were 31 variables in the deboning data, and we divided into them two categories: data obtained before and after deboning. Discriminant function analysis was then applied into the data to determined the accuracy of gender discrimination in Hanwoo. The result showed that Hanwoo could be classified by gender with 99.2% of accuracy when using all 31 variables. In detail, it was possible to identify 93 of 94 bulls (98.9%), 96 of 96 cows (100%) and 74 of 75 steers (98.7%). The most significant variables was chuck, sirloin, armbone shin, plates, retail and cuts percentage, sequentially. With variables obtainable before deboning, accuracies of classification were 91.5% for bulls, 92.7% for cows, and 89.3% for steers. The most significant variables was water, cold carcass weight and back-fat thickness. The discrimination accuracy was higher with data obtainable after deboning: bulls (98.9%), cows (99.0%) and steers (98.7%). In this case, chuck, sirloin and armbone shin were the factors determined the classification ability. This study showed that Hanwoo can be classified based on deboning data with appropriate statistics, further suggesting weight of cut of beef might be the standard for gender classification.

• 대한산업공학회지
• /
• 제28권2호
• /
• pp.223-231
• /
• 2002

Recently, work-related musculoskeletal disorders(WMSDs) have rapidly increased and have been a major issue in the field of industrial safety. Of several physical risk factors for WMSDs, which include postures, vibration, repetitive work, speed or acceleration of movements, etc., awkward postures have been known as one of the major causes of WMSDs. For reducing the potential for injury as a result of postures, cost effective quantification of the magnitude for physical exposure to poor working postures is important and needed. To do this, several postural classification schemes have been developed and used in industrial sites. It is known that perceived discomfort for joint motions and muscle strength for females were much less than those for males. However, the existing postural classification schemes were developed without considering these gender effects. This study aims to develop a new postural classification scheme for female workers, based on the perceived discomfort for joint motions. The result showed that there was significant difference between the schemes for female and male. It was also found that when compared with OWAS, RULA and REBA, postural load was quantified more precisely with the developed scheme. It is recommended that different schemes according to gender of workers involved in work be used in order to accurately evaluate postural load of work postures.

• 한국전자통신학회논문지
• /
• 제13권1호
• /
• pp.169-174
• /
• 2018

시민들의 안전을 위한 영상통합관제센터에는 수많은 CCTV 카메라가 연결되어 많은 채널의 영상을 소수의 관제사가 관제하는데 어려움이 있다. 본 논문에서는 많은 채널의 영상을 효과적으로 관제하기 위하여 안전지도와 연계한 지능형 영상보안 시스템을 제안한다. 안전지도는 범죄 발생 빈도를 데이터베이스로 구축하고, 범죄 발생 위험 정도를 표현하고, 범죄 취약 계층인 여성이 범죄 위험 지역으로 진입하면 영상통합관제센터의 관제사가 주목할 수 있도록 한다. 성별 구분을 보행자 검출 및 추적 그리고 딥러닝을 통하여 성별을 구분한다. 보행자 검출은 Adaboost 알고리즘을 이용하고, 보행자 추적을 위한 확률적 데이터 연관 필터(probablistic data association filter)를 적용한다. 보행자의 성별을 구분하기 위하여 비교적 간단한 AlexNet를 적용하여 성별을 판별한다. 실험을 통하여 제안하는 성별 구분 방법이 종래의 알고리즘에 비하여 성별 구분에 효과적임을 보인다. 또한 안전지도와 연계한 지능형 영상보안 시스템 구현 결과를 소개한다.

• Imaging Science in Dentistry
• /
• 제47권3호
• /
• pp.149-155
• /
• 2017

Purpose: The classification of impacted maxillary third molars(IMTMs) facilitates interdisciplinary communication and helps estimate the degree of surgical difficulty. Thus, this study aimed to develop a 3-dimensional classification of the position and depth of impaction of IMTMs and to estimate their prevalence with respect to gender and age. Materials and Methods: This cross-sectional retrospective study analyzed images in sagittal and coronal cone-beam computed tomography (CBCT) sections of 300 maxillary third molars. The proposed classification was based on 3 criteria: buccolingual position (buccal, lingual, or central), mesial-distal position (mesioangular, vertical, or distoangular), and depth of impaction (low, medium, or high). CBCT images of IMTMs were classified, and the associations of the classifications with gender and age were examined using analysis of variance with the Scheffe post-hoc test. To determine the associations among the 3 classifications, the chi-square test was used (P<.05). Results: No significant association of the classifications with gender was observed. Age showed a significant relationship with depth of impaction (P=.0001) and mesial-distal position (P=.005). The most common positions were buccal(n=222), vertical(n=184), and low (n=124). Significant associations among the 3 tested classifications were observed. Conclusion: CBCT enabled the evaluation of IMTMs in a 3-dimensional format, and we developed a proposal for a new classification of the position and depth of impaction of IMTMs.

• 한국임상약학회지
• /
• 제23권2호
• /
• pp.114-122
• /
• 2013

Background: Drugs should be evaluated in appropriate subjects representing potential population to take the drugs. This study focuses on gender factor and aims to make known the appropriateness of considering gender difference on clinical evaluation of drug with domestic data related to drug use. Methods: To understand gender difference shown in drug use, three types of domestic statistical data (prevalence of chronic disease, number of outpatient with major concerning disease, and consumption of medicine) were analyzed and compared according to gender. Results: Three of fifteen chronic diseases which were analyzed, showed significantly higher prevalence in women than in men, and three were vice versa. Meanwhile, the sex ratio of outpatients was significantly different in 22 major concerning diseases. Among the drug groups coded by Anatomical Therapeutic Chemical (ATC) Classification System, the consumption of most drug groups was generally higher in women than in men except for one group coded G (genito-urinary system and sex hormones). Conclusion: Gender difference should be considered in domestic clinical evaluation of drug and domestic guidance for reflecting gender difference should be established.