http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
얼굴 영상 기반의 심박수 추정을 위한 딥러닝 모델의 경량화 기법
황규태,박명근,이상준 대한임베디드공학회 2023 대한임베디드공학회논문지 Vol.18 No.2
This paper proposes a deep learning method for estimating the heart rate from facial videos. Our proposed method estimates remote photoplethysmography (rPPG) signals to predict the heart rate. Although there have been proposed several methods for estimating rPPG signals, most previous methods can not be utilized in low-power single board computers due to their computational complexity. To address this problem, we construct a lightweight student model and employ a knowledge distillation technique to reduce the performance degradation of a deeper network model. The teacher model consists of 795k parameters, whereas the student model only contains 24k parameters, and therefore, the inference time was reduced with the factor of 10. By distilling the knowledge of the intermediate feature maps of the teacher model, we improved the accuracy of the student model for estimating the heart rate. Experiments were conducted on the UBFC-rPPG dataset to demonstrate the effectiveness of the proposed method. Moreover, we collected our own dataset to verify the accuracy and processing time of the proposed method on a real-world dataset. Experimental results on a NVIDIA Jetson Nano board demonstrate that our proposed method can infer the heart rate in real time with the mean absolute error of 2.5183 bpm.
황규태,김수석 한국조명전기설비학회 1992 조명.전기설비 Vol.6 No.1
In this paper, the basic idea of optimum electric system design by means of the forecasting of maximum demand is presented, and the load characteristics and practical operating conditions are based on the technical data. After reconstruction of th model plant by use of above method, power supply reliability, future extention, initial cost, and running cost saving effects are analyzed. As a result, it is verified that the systems wherein the power is supply to each load frm main transformer whose capacity is calculated by forecasting are economic rather than the systems wherein the power is supply to each electric feeders from each corresponding transformer.
경전자 도달법을 이용한 일차성 인공 고관절 전치환술 후 고관절 탈구 발생의 위험 요인
황규태 ( Kyu Tae Hwang ),김영호 ( Young Ho Kim ),김이석 ( Yee Suk Kim ),봉현종 ( Hyun Jong Bong ),최일용 ( Il Yong Choi ) 대한고관절학회 2010 Hip and Pelvis Vol.22 No.1
Purpose: We wanted to evaluate the risk factors that predispose a patient to dislocation after undergoing primary total hip arthroplasty with the transtrochanteric approach. Materials and Methods: Between July 1995 and May 2007, 593 consecutive total hip arthroplasties using the trantrochanteric approach were performed. A matched comparative study was performed for the dislocated group(18 hips) and the non-dislocated control group(18 hips). The patient-related factors and mechanical factors were retrospectively reviewed to evaluate the risk factors for dislocation. Results: Dislocation occurred in 18 hips(3.04%). The mean age was 57.1 years in the dislocated group and 55.2 years for all of the patients (p>0.05). Statistical analyses of the BMI, inclination and anteversion of the cup, lowering of the hip center, a leg length discrepancy, the size of the femoral head and nonunion of the greater trochanter revealed no significant differences between the two groups. Nonunion of the greater trochanter was observed in 16 hips(2.84%). The risk of dislocation was 8.5 times higher in the patients with excessive alcohol intake (p<0.05). The combination of more than 3 risk factors significantly affected the incidence of dislocation after total hip arthroplasty (p<0.05). Conclusion: After primary total hip arthroplasty with the transtrochanteric approach, in the cohort of this study, excessive alcohol intake was the main risk factor for dislocation and a combination of risk factors contributed to dislocation.