• Journal of Information Processing Systems
• /
• 제18권3호
• /
• pp.311-318
• /
• 2022

Tobacco sales enterprises often need to summarize and verify the daily sales bills, which may consume substantial manpower, and manual verification is prone to occasional errors. The use of artificial intelligence technology to realize the automatic identification and verification of such bills offers important practical significance. This study presents a novel multi-branch residual network for tobacco sales bills to improve the efficiency and accuracy of tobacco sales. First, geometric correction and edge alignment were performed on the input sales bill image. Second, the multi-branch residual network recognition model is established and trained using the preprocessed data. The comparative experimental results demonstrated that the correct recognition rate of the proposed method reached 98.84% on the China Tobacco Bill Image dataset, which is superior to that of most existing recognition methods.

• 스마트미디어저널
• /
• 제9권2호
• /
• pp.9-15
• /
• 2020

본 논문에서는 새롭게 제안하는 Multi-Path Encoder-Decoder 의 구조를 바탕으로 두개의 가지로 구성된 심층신경망을 통해서 영상 이미지에서 물체를 하나의 객체 단위로 분할 검출하는 방법을 제안하였다. 각 가지는 중심점 검출 가지(Dot branch), 객체 분할 가지(Segmentation branch)라 하고 중심점 검출 가지는 이미지로부터 각 객체의 중심점을 찾는 역할을 수행하고, 객체 분할 가지는 각 객체의 영역을 이미지로부터 분할하는 역할을 수행한다. 실험에서는 CVPPP 식물 이미지의 나뭇잎을 각각 구분하도록 학습 하였으며 중심점 검출 가지는 각 나뭇잎의 중심점들을 찾아내고, 객체 분할 가지는 원본 이미지와 찾아낸 중심점 이미지를 통하여 각 중심점에 해당하는 나뭇잎의 픽셀 분할 영역을 최종적으로 예측하게 된다. 기존의 객체 분할에서는 다양한 크기, 위치의 앵커박스를 만들어서 많은 영역(N > 1k)의 물체를 확인해야하는 연산량 문제점 혹은 이미지에서 고정되지 않는 총 객체의 개수를 예측하기 어려웠던 문제가 있었다. 제안한 심층신경망에서는 중심점을 기반으로 객체를 찾아내는 효과적인 방법을 제안하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권5호
• /
• pp.2346-2361
• /
• 2017

A major challenge in cloud infrastructure is the efficient allocation of virtual network elements on top of substrate network elements. Path algebra is a mathematical framework which allows the validation and convergence analysis of the mono-constraint or multi-constraint routing problems independently of the network topology or size. The present study proposes a new heuristic approach based on mathematical framework "paths algebra" to map virtual nodes and links to substrate nodes and paths in cloud. In this approach, we define a measure criterion to rank the substrate nodes, and map the virtual nodes to substrate nodes according to their ranks by using a greedy algorithm. In addition, considering multi-constraint routing in virtual link mapping stage, the used paths algebra framework allows a more flexible and extendable embedding. Obtained results of simulations show appropriate improvement in acceptance ratio of virtual networks and cost incurred by the infrastructure networks.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권9호
• /
• pp.2483-2504
• /
• 2023

Most existing low-light enhancement algorithms either use a large number of training parameters or lack generalization to real-world scenarios. This paper presents a novel lightweight and robust pixel-wise polynomial approximation-based deep network for low-light image enhancement. For mapping the low-light image to the enhanced image, pixel-wise higher-order polynomials are employed. A deep convolution network is used to estimate the coefficients of these higher-order polynomials. The proposed network uses multiple branches to estimate pixel values based on different receptive fields. With a smaller receptive field, the first branch enhanced local features, the second and third branches focused on medium-level features, and the last branch enhanced global features. The low-light image is downsampled by the factor of 2b-1 (b is the branch number) and fed as input to each branch. After combining the outputs of each branch, the final enhanced image is obtained. A comprehensive evaluation of our proposed network on six publicly available no-reference test datasets shows that it outperforms state-of-the-art methods on both quantitative and qualitative measures.

• Journal of Electrical Engineering and Technology
• /
• 제10권4호
• /
• pp.1752-1758
• /
• 2015

This paper presents partial discharge localization in stator winding of generators using multi-conductor transmission line (MTL) and RLC ladder network models. The high-voltage (HV) winding of a 6kV/250kW generator has been modeled by MATLAB software. The simulation results of the MTL and the RLC ladder network models have been evaluated with the measurements results in the frequency domain by applying of the Pearson’s correlation coefficients. Two PD generated calibrator signals in kHz and MHz frequency range were injected into different points of generator winding and the signals simulated/measured at the both ends of the winding. For partial discharge localization in stator winding of generators is necessary to calculate the frequency spectrum of the PD current signals and then estimate the poles of the system from the calculated frequency spectrum. Finally, the location of PD can be estimated. This theory applied for the above generator and the simulation/measured results show the good correlation for PD Location for RLC ladder network and MTL models in the frequency range of kHz (10kHz<f<1MHz) and MHz (1MHz<f<5MHz) respectively.

• 한국해양공학회지
• /
• 제11권4호
• /
• pp.152-158
• /
• 1997

An analyzing method for pressure fluctuations in oil hydraulic pipe network was developed in this study. The object pipe network has multi-branch configuration, and the pipelines of it are composed of steel tubes, flexible hoses. Also, accumulators, orifices and lumped oil volume components are attached on it. Transfer matrix method, in other words impedance method, was used for the analysis. The reliability and usefulness of the analyzing method were confirmed by investigation computed results and experimental results got in this study.

• Earthquakes and Structures
• /
• 제15권4호
• /
• pp.431-441
• /
• 2018

This study provides fragility-based assessment of seismic performance of reinforced concrete bridges. Seismic fragility curves were created using nonlinear analysis (NA) and artificial neural networks (ANNs). Nonlinear response history analyses were performed, in order to calculate the seismic performances of the bridges. To this end, 306 bridge-earthquake cases were considered. A multi-layered perceptron (MLP) neural network was implemented to predict the seismic performances of the selected bridges. The MLP neural networks considered herein consist of an input layer with four input vectors; two hidden layers and an output vector. In order to train ANNs, 70% of the numerical results were selected, and the remained 30% were employed for testing the reliability and validation of ANNs. Several structures of MLP neural networks were examined in order to obtain suitable neural networks. After achieving the most proper structure of neural network, it was used for generating new data. A total number of 600 new bridge-earthquake cases were generated based on neural simulation. Finally, probabilistic seismic safety analyses were conducted. Herein, fragility curves were developed using numerical results, neural predictions and the combination of numerical and neural data. Results of this study revealed that ANNs are suitable tools for predicting seismic performances of RC bridges. It was also shown that yield stresses of the reinforcements is one of the important sources of uncertainty in fragility analysis of RC bridges.

• 대한원격탐사학회지
• /
• 제39권2호
• /
• pp.143-156
• /
• 2023

인공위성이 촬영한 영상의 내용을 정확하게 분석하기 위해서는 영상에 존재하는 구름 영역을 정확하게 인지하는 것이 필요하다. 최근 다양한 분야에서 딥러닝(deep learning) 모델이 뛰어난 성능을 보여줌에 따라 구름 영역 검출을 위해 딥러닝 모델을 적용한 방법들이 많이 제안되고 있다. 하지만 현재 구름 영역 검출 방법들은 의미 영역 분할 방법의 네트워크 구조를 그대로 사용하여 구름 검출 성능을 향상하는 데는 한계가 있다. 따라서 본 논문에서는 구름 검출 데이터 세트에 다중 브랜치 네트워크 구조 탐색을 적용하여 구름 영역 검출에 최적화된 네트워크 모델을 생성함으로써 구름 검출 성능을 향상하는 방법을 제안한다. 또한 구름 검출 성능을 향상하기 위하여 의미 영역 분할 모델의 학습 단계와 평가 단계의 평가 기준 불일치를 해소하기 위해 제안된 soft intersection over union (IoU) 손실 함수를 사용하고, 다양한 데이터 증강 방법을 적용하여 학습 데이터를 증가시켰다. 본 논문에서 제안된 방법의 성능을 검증하기 위하여 아리랑위성 3/3A호에서 촬영한 영상으로 구성된 구름 검출 데이터 세트를 사용하였다. 먼저 제안 방법과 의미 영역 분할 데이터 세트에서 탐색된 기존 네트워크 모델의 성능을 비교하였다. 실험 결과, 제안 방법의 mean IoU는 68.5%이며, 기존 모델보다 mIoU 측면에서 4%의 높은 성능을 보여주었다. 또한 soft IoU 손실 함수를 포함한 다섯 개의 손실 함수를 적용하여 손실 함수에 따른 구름 검출 성능을 분석하였으며, 실험 결과 본 연구에서 사용한 soft IoU 함수가 가장 좋은 성능을 보여주었다. 마지막으로 의미 영역 분할 분야에서 활용되는 최신 네트워크 모델과 제안 방법의 구름 검출 성능을 비교하였다. 실험 결과, 제안 모델이 의미 영역 분할 분야의 최신 모델들보다 mIoU와 정확도 측면에서 더 나은 성능을 보여주는 것을 확인하였다.

• 한국정보통신학회논문지
• /
• 제26권10호
• /
• pp.1551-1558
• /
• 2022

본 논문에서는 FEC (Fog/Edge Computing) 환경에서 다중 분기구조의 부분 오프로딩을 위해 모바일 장치와 에지서버로 구성된 2계층 협력 컴퓨팅 시스템을 제안한다. 제안 시스템은 다중 분기구조에 대한 재구성 선형화 기법을 적용하여 응용 서비스 처리를 분할하는 알고리즘과 모바일 장치와 에지 서버 간의 부분 오프로딩을 통한 최적의 협업 알고리즘을 포함한다. 또한 계산 오프로딩 및 CNN 계층 스케줄링을 지연시간 최소화 문제로 공식화하고 시뮬레이션을 통해 제안 시스템의 효과를 분석한다. 실험 결과 제안 알고리즘은 DAG 및 체인 토폴로지 모두에 적합하고 다양한 네트워크 조건에 잘 적응할 수 있으며, 로컬이나 에지 전용 실행과 비교하여 효율적인 작업 처리 전략 및 처리시간을 제공한다. 또한 제안 시스템은 모바일 장치에서의 응용 서비스 최적 실행을 위한 모델의 경량화 및 에지 리소스 워크로드의 효율적 분배 관련 연구에 적용 가능하다.

• Industrial Engineering and Management Systems
• /
• 제14권4호
• /
• pp.347-371
• /
• 2015

A hybrid procedure based on multi-Criteria Satisfaction Analysis (MUSA) and a Network Data Envelopment Analysis (NDEA) is proposed to evaluate the relative efficiency of customer services in bank branches. First, a three-stage process including sub-processes such as customer expectations, customer satisfaction, and customer loyalty, is defined to model the banking customer services. Then, fulfillment of customer expectations, customer loyalty level, and the customer satisfaction degree are measured and quantified through a multi-dimensional questionnaire based on customers' perceptions analysis and MUSA method, respectively. The customer services scores and the other criteria such as mean of employee evaluation score, operation costs, assets, deposits, loans, number of accounts are considered in network three-stage DEA model. The proposed NDEA model is formed based on multipliers perspective, output-oriented, and constant return to scale assumptions. The proposed NDEA model quantifies and assesses the total efficiency of main process and assigns the efficiency to customer expectations, customer satisfactions, and customer loyalties sub-processes in bank branches. The whole procedure is applied on 30 bank branches in IRAN. The proposed approach can be used in other organizations such as airports, airline agencies, urban transportation systems, railway organizations, chain stores, chain restaurants, public libraries, and entertainment centers.