• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권10호
• /
• pp.4930-4951
• /
• 2018

The deep learning based speech enhancement has shown considerable success. However, it still suffers performance degradation under mismatch conditions. In this paper, an adaptation method is proposed to improve the performance under noise mismatch conditions. Firstly, we advise a noise aware training by supplying identity vectors (i-vectors) as parallel input features to adapt deep neural network (DNN) acoustic models with the target noise. Secondly, given a small amount of adaptation data, the noise-dependent DNN is obtained by using $L_2$ regularization from a noise-independent DNN, and forcing the estimated masks to be close to the unadapted condition. Finally, experiments were carried out on different noise and SNR conditions, and the proposed method has achieved significantly 0.1%-9.6% benefits of STOI, and provided consistent improvement in PESQ and segSNR against the baseline systems.

• 한국융합학회논문지
• /
• 제9권5호
• /
• pp.1-6
• /
• 2018

스마트그리드에서 정확한 단기 부하 예측을 통한 자원의 이용 계획은 에너지 시스템 운영의 불확실성을 줄이고 운영 효율을 높이는데 있어서 매우 중요하다. 단기 부하 예측에 얕은 신경회로망을 포함한 다수의 머신 러닝 기법이 적용되어왔지만 예측 정확도의 개선이 요구되고 있다. 최근에는 컴퓨터 비전이나 음성인식 분야에서 심층 신경회로망의 뛰어난 연구 결과로 인해 심층 신경회로망을 단기 전력수요 예측에 적용해 예측 정확도를 개선하려는 시도가 주목 받고 있다. 본 논문에서는 일별 전력 부하 첨두치를 예측하기 위한 다층신경회로망 구조의 심층 신경회로망 모델을 제안한다. 제안된 심층 신경회로망은 층별 학습이 선행된 후 전체 모델의 학습이 이루어진다. 한국전력거래소에서 얻은 4년 동안의 일별 전력 수요 데이터를 사용, 하루 및 이틀 앞선 전력수요 첨두치를 예측하는 심층 신경회로망 모델을 구축하고 예측 정확도를 비교, 평가한다.

• 대한의용생체공학회:의공학회지
• /
• 제40권2호
• /
• pp.48-54
• /
• 2019

This paper proposed a method of validation data augmentation for improving the grading accuracy of diabetic macular edema (DME) using deep learning. The data augmentation technique is basically applied in order to secure diversity of data by transforming one image to several images through random translation, rotation, scaling and reflection in preparation of input data of the deep neural network (DNN). In this paper, we apply this technique in the validation process of the trained DNN, and improve the grading accuracy by combining the classification results of the augmented images. To verify the effectiveness, 1,200 retinal images of Messidor dataset was divided into training and validation data at the ratio 7:3. By applying random augmentation to 359 validation data, $1.61{\pm}0.55%$ accuracy improvement was achieved in the case of six times augmentation (N=6). This simple method has shown that the accuracy can be improved in the N range from 2 to 6 with the correlation coefficient of 0.5667. Therefore, it is expected to help improve the diagnostic accuracy of DME with the grading information provided by the proposed DNN.

• 한국음향학회지
• /
• 제36권6호
• /
• pp.401-406
• /
• 2017

본 논문에서는 양서류 울음소리를 통한 종 인식 시스템 개발을 위해, 음향 신호 분석에서 활용되는 주요 알고리즘의 인식 성능을 평가했다. 먼저, 멸종위기 종을 포함하여 총 9 종의 양서류를 선정하여, 각 종별 울음소리를 야생에서 녹음하여 실험 데이터를 구축했다. 성능평가를 위해, MFCC(Mel Frequency Cepstral Coefficient), RCGCC(Robust Compressive Gammachirp filterbank Cepstral Coefficient), SPCC(Subspace Projection Cepstral Coefficient)의 세 특징벡터와 GMM(Gaussian Mixture Model), SVM(Support Vector Machine), DBN-DNN(Deep Belief Network - Deep Neural Network)의 세 인식기가 고려됐다. 추가적으로, 화자 인식에 널리 사용되는 i-vector를 이용한 인식 실험도 수행했다. 인식 실험 결과, SPCC-SVM의 경우 98.81 %로 가장 높은 인식률을 확인 할 수 있었으며, 다른 알고리즘에서도 90 %에 가까운 인식률을 확인했다.

• 한국멀티미디어학회논문지
• /
• 제22권11호
• /
• pp.1300-1312
• /
• 2019

BSTRACT In this paper, PM10 forecast model using DNN(Deep Neural Network) is developed for Seoul region. The previous Julian forecast model has been developed using weather and air quality data of Seoul region only. This model gives excellent results for accuracy and false alarm rates, but poor result for POD(Probability of Detection). To solve this problem, an WA(Wide Area) forecasting model that uses Chinese data is developed. The data is highly correlated with the emergence of high concentrations of PM10 in Korea. As a result, the WA model shows better accuracy, and POD improving of 3%(D+0), 21%(D+1), and 36%(D+2) for each forecast period compared with the Julian model.

• Structural Engineering and Mechanics
• /
• 제85권1호
• /
• pp.15-27
• /
• 2023

Using artificial intelligence and internet of things methods in engineering and industrial problems has become a widespread method in recent years. The low computational costs and high accuracy without the need to engage human resources in comparison to engineering demands are the main advantages of artificial intelligence. In the present paper, a deep neural network (DNN) with a specific method of optimization is utilize to predict fundamental natural frequency of a cylindrical structure. To provide data for training the DNN, a detailed numerical analysis is presented with the aid of functionally modified couple stress theory (FMCS) and first-order shear deformation theory (FSDT). The governing equations obtained using Hamilton's principle, are further solved engaging generalized differential quadrature method. The results of the numerical solution are utilized to train and test the DNN model. The results are validated at the first step and a comprehensive parametric results are presented thereafter. The results show the high accuracy of the DNN results and effects of different geometrical, modeling and material parameters in the natural frequencies of the structure.

• 전력전자학회:학술대회논문집
• /
• 전력전자학회 2019년도 추계학술대회
• /
• pp.70-72
• /
• 2019

For the safe and reliable operation of Lithium-ion batteries in Electric Vehicles (EVs) or Energy Storage Systems (ESSs), it is essential to have accurate information of the battery such as State of Charge (SOC). Many kinds of different techniques to estimate the SOC of the batteries have been developed so far such as the Kalman Filter. However, when it is applied to the multiple number of batteries it is difficult to maintain the accuracy of the estimation over all cells due to the difference in parameter value of each cell. Moreover the difference in the parameter of each cell may become larger as the operation time accumulates due to aging. In this paper a novel Deep Neural Network (DNN) based SOC estimation method for multi cell application is proposed. In the proposed method DNN is implemented to learn non-linear relationship of the voltage and current of the lithium-ion battery at different SOCs and different temperatures. In the training the voltage and current data of the Lithium battery at charge and discharge cycles obtained at different temperatures are used. After the comprehensive training with the data obtained with a cell resulting estimation algorithm is applied to the other cells. The experimental results show that the Mean Absolute Error (MAE) of the estimation is 0.56% at 25℃, and 3.16% at 60℃ with the proposed SOC estimation algorithm.

• Geomechanics and Engineering
• /
• 제28권5호
• /
• pp.477-491
• /
• 2022

In the present study, the effects of thermal loading on the buckling and resonance frequency of graphene platelets (GPL) reinforced nano-composites are examined. Functionally graded (FG) material properties are considered in thickness direction for the thermal responses of the composite. The equivalent material properties are obtained using Halphin-Tsai nano-mechanical model for composite layers. Moreover, the effects of nano-scale sizes are taken into account, employing functionally modified couple stress (FMCS) parameter. In this regard, for the first time, it is demonstrated that at certain values of GPL weight fraction, thermal buckling occurs. In obtaining results of vibrational behavior, both analytical solution and deep neural network (DNN) methods are used. The DNN method needs low computational costs to predict the resonance behavior. A comprehensive parametric study is conducted to indicate the effects of several geometrical, material, and loading conditions on the vibrational and buckling behavior of cylindrical shell structures made of GPL-nanocomposites. It is shown that the effect of temperature change on the occurrence of buckling is vital while it has a negligible impact on the resonance frequency of the structure. Moreover, the size-dependency of the results is demonstrated, and it cannot be neglected in nano-scales.

• 한국건축시공학회:학술대회논문집
• /
• 한국건축시공학회 2022년도 가을 학술논문 발표대회
• /
• pp.223-224
• /
• 2022

The occurrence of defect, one of the major risk elements, gives rise to construction delays and additional costs. Although construction companies generally prefer to use a method of identifying and classifying the causes of defects, a system for predicting the rise of defects becomes important matter to reduce this harmful issue. However, the currently used methods are kinds of reactive systems that are focused on the defects which occurred already, and there are few studies on the occurrence of defects with prediction systems. This paper is about preliminary study on the development of judgemental algorithm that informs us whether additional works related to defect issue are needed or not. Among machine learning techniques, deep neural network was utilized as prediction model which is a major component of algorithm. It is the most suitable model to be applied to the algorithm when there are 8 hidden layers and the average number of nodes in each hidden layer is 70. Ultimately, the algorithm can identify and defects that may arise in later and contribute to minimize defect frequency.

• 한국건축시공학회지
• /
• 제21권6호
• /
• pp.665-676
• /
• 2021

노후 건축물의 수가 증가함에 따라, 건물의 안전진단, 유지 보수에 대한 중요성이 증가하고 있다. 기존 외관 조사는 점검자의 주관적인 판단이 수반되어 평가 결과가 다르고 객관성과 신뢰성이 떨어진다. 따라서 본 연구는 기존 연구를 통해 기실시된 외관 조사 및 상태 평가 프로세스의 한계를 제시하였으며, UAV, Laser Scanner를 통해 3D Point Cloud 데이터를 수집하였다. 또한, Reverse Engineering 기술을 이용하여 3D 모델을 생성한 후 객관적인 상태평가 데이터를 취득하였다. 이후 기존의 정밀검사 데이터와 정밀 안전진단 데이터를 활용하여 DNN 구조를 생성하고, 고정밀도 측정 장치를 이용하여 얻은 상태평가 데이터를 적용하여 객관적인 건물안전등급을 산출하였다. 자동화된 프로세스는 20개의 노후된 건축물에 적용되며 동일 면적 건축물 기준 수작업으로 실시되는 안전진단의 시간에 비해 약 50% 감소하였다. 이후 본 연구에서는 안전등급 결과값과 기존값을 비교하여 안전등급 산출과정의 정확성을 검증하고 약 90%의 높은 정확도를 가진 DNN을 구축하였다. 이는 향후 노후 건물의 안전등급 산정의 신뢰성이 향상되고 비용과 시간을 절약해 경제성이 향상될 것으로 기대된다.