• Journal of Intelligence and Information Systems
• /
• v.26 no.2
• /
• pp.1-25
• /
• 2020

In this paper, we suggest an application system architecture which provides accurate, fast and efficient automatic gasometer reading function. The system captures gasometer image using mobile device camera, transmits the image to a cloud server on top of private LTE network, and analyzes the image to extract character information of device ID and gas usage amount by selective optical character recognition based on deep learning technology. In general, there are many types of character in an image and optical character recognition technology extracts all character information in an image. But some applications need to ignore non-of-interest types of character and only have to focus on some specific types of characters. For an example of the application, automatic gasometer reading system only need to extract device ID and gas usage amount character information from gasometer images to send bill to users. Non-of-interest character strings, such as device type, manufacturer, manufacturing date, specification and etc., are not valuable information to the application. Thus, the application have to analyze point of interest region and specific types of characters to extract valuable information only. We adopted CNN (Convolutional Neural Network) based object detection and CRNN (Convolutional Recurrent Neural Network) technology for selective optical character recognition which only analyze point of interest region for selective character information extraction. We build up 3 neural networks for the application system. The first is a convolutional neural network which detects point of interest region of gas usage amount and device ID information character strings, the second is another convolutional neural network which transforms spatial information of point of interest region to spatial sequential feature vectors, and the third is bi-directional long short term memory network which converts spatial sequential information to character strings using time-series analysis mapping from feature vectors to character strings. In this research, point of interest character strings are device ID and gas usage amount. Device ID consists of 12 arabic character strings and gas usage amount consists of 4 ~ 5 arabic character strings. All system components are implemented in Amazon Web Service Cloud with Intel Zeon E5-2686 v4 CPU and NVidia TESLA V100 GPU. The system architecture adopts master-lave processing structure for efficient and fast parallel processing coping with about 700,000 requests per day. Mobile device captures gasometer image and transmits to master process in AWS cloud. Master process runs on Intel Zeon CPU and pushes reading request from mobile device to an input queue with FIFO (First In First Out) structure. Slave process consists of 3 types of deep neural networks which conduct character recognition process and runs on NVidia GPU module. Slave process is always polling the input queue to get recognition request. If there are some requests from master process in the input queue, slave process converts the image in the input queue to device ID character string, gas usage amount character string and position information of the strings, returns the information to output queue, and switch to idle mode to poll the input queue. Master process gets final information form the output queue and delivers the information to the mobile device. We used total 27,120 gasometer images for training, validation and testing of 3 types of deep neural network. 22,985 images were used for training and validation, 4,135 images were used for testing. We randomly splitted 22,985 images with 8:2 ratio for training and validation respectively for each training epoch. 4,135 test image were categorized into 5 types (Normal, noise, reflex, scale and slant). Normal data is clean image data, noise means image with noise signal, relfex means image with light reflection in gasometer region, scale means images with small object size due to long-distance capturing and slant means images which is not horizontally flat. Final character string recognition accuracies for device ID and gas usage amount of normal data are 0.960 and 0.864 respectively.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.559-560
• /
• 2018

순환 신경망(RNN) 기반의 Long Short-Term Memory(LSTM)는 자연어처리 분야에서 우수한 성능을 보이는 모델이다. 음성을 문자로 변환해주는 Speech to Text (STT)를 이용해 자막을 생성하고, 생성된 자막을 다른 언어로 동시에 번역을 해주는 서비스가 활발히 진행되고 있다. STT를 사용하여 자막을 추출하는 경우에는 마침표가 없이 전부 연결된 문장이 생성되기 때문에 정확한 번역이 불가능하다. 본 논문에서는 영어자막의 자동 번역 시, 정확도를 높이기 위해 텍스트를 문장으로 분할하여 마침표를 생성해주는 방법을 제안한다. 이 때, LSTM을 이용하여 데이터를 학습시킨 후 테스트한 결과 62.3%의 정확도로 마침표의 위치를 예측했다.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1997.04a
• /
• pp.150-155
• /
• 1997

In this paper,we present neural network for control of Left Ventricular Assist Device(LVAD)system with a pneumatically driven mock cirulation system. It is necessary to apply high perfomance control techniques, since the LVAD system represent nonlinear and time-varing characteristics. Fortunately, the neural network can be applied to control of a nonliner dynamic system by learning capability. In this study,we identify the LVAD system with neural network and control the LVAD system by PID controller and neural network feedforward controller. The ability and effectiveness of controlling the LVAD system using the proposed algorithm will be demonstrated by computer simulation and experiment.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2001.06a
• /
• pp.121-124
• /
• 2001

In industrial actual control system, PID controller has been used with its high delicate control system in position control system. PID controller has simple structure and superior ability in several characteristics. When the response of system is changed by delay time, variable load , disturbances and external environment, control gain of PID controller must be readjusted on the system dynamic characteristics. Therefore, a control ability of PID controller is degraded when th control gain is inappropriately determined. When the response characteristic of system is changed under a condition, control gain of PID controller must be changed adaptively to be a waited response of system. In this paper an adaptive-tuning type PID controller is constructed by self-recurrent Neural Network(SRNN). applying back-propagation(BP) algorithm. Form the result of computer simulation in the proposed controller, its usefulness is verified.

• Journal of the Korean Society of Visualization
• /
• v.19 no.1
• /
• pp.57-61
• /
• 2021

In order to determine erythrocyte sedimentation rate (ESR) indicating acute phase inflammation, a Westergren method has been widely used because it is cheap and easy to be implemented. However, the Westergren method requires quite a long time for 1 hour. In this study, a gated recurrent unit (GRU) neural network was used to reduce measurement time of ESR evaluation. The sedimentation sequences of the erythrocytes were acquired by the camera and data processed through image processing were used as an input data into the neural network models. The performance of a proposed models was evaluated based on mean absolute error. The results show that GRU model provides best accurate prediction than others within 30 minutes.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.11a
• /
• pp.1025-1027
• /
• 2020

태권도 경기와 같이 동작의 정확한 기술을 판별하여 유효득점화하는 시스템에서는 점수 체계의 정확성과 전문성이 필요하다. 기존에 시행되었던 심판 판정은 객관성과 신뢰성의 결여 문제가 존재하여 이를 대체하기 위한 방법으로 전자호구가 도입되었다. 하지만 전자호구는 타격 강도에 따라 분류되는 문제로 인해 태권도 기술이 아닌 변칙 발차기 기술에서도 유효득점이 처리되는 문제가 발생하였다. 본 논문에서는 변칙 발차기와 일반 발차기를 분류하여 변칙 발차기에서의 유효득점을 무효 득점화 시키기 위한 분류 모델을 제안하였다. 순환 신경망 모델인 LSTM을 이용하여 변칙 발차기와 일반 발차기를 분류하였으며 94.90%의 정확도를 보였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.1203-1206
• /
• 2021

소음은 난청, 스트레스 등의 원인이 된다. 본 연구에서는 ANC(Active Noise Cancellation)을 바탕으로, 기술적인 방법을 통해 소음을 저감 시키는 스피커를 구현하였다. ANC 란 소음 주파수의 위상을 180° 변환하여 주파수와 레벨이 동일한 역 소음을 발생시켜 주변 소음을 저감, 차단하는 기술이다. 현재 시중 제품들에 적용되는 일반적인 ANC 의 경우, 피드백(Feedback) 방식이라는 점과 시간 지연(Time gap)이 발생한다는 한계가 있다. 이를 보완하기 위해 AI 학습으로 소음을 미리 예측하여 시간 지연을 줄이는 방법을 고안했다. 순환 신경망(RNN)의 장기의존성 문제를 해결하는 시계열 예측 딥러닝 알고리즘인 LSTM(Long Short-Term Memory Network) 모델을 사용하였다. 또한, AI 학습 효율을 향상시킬 수 있는 하드웨어 장비들을 활용하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.05a
• /
• pp.549-551
• /
• 2023

차량용 인포테인먼트 시스템은 차량 내부에서 정보와 엔터테인먼트 기능을 제공하는 시스템으로, 현재 급격한 성장세를 보이고 있다. 이에 따라 많은 기업이 차량용 인포테인먼트 관련 기술을 연구하고 개발하고 있다. 이는 결국 차량에서 발생하는 트래픽이 이전보다 증가하는 것을 의미한다. 차량 발생 트래픽은 모바일 트래픽과 달리 시간에 따라 뚜렷한 발생 패턴을 보인다. 이러한 특성을 고려하여 RNN, LSTM, GRU 세 가지 종류의 순환 신경망 모델을 활용하여 차량 트래픽 예측 모델을 구현하였고 시간대별 모델 성능을 비교한 결과, LSTM이 가장 우수한 성능을 보였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.05a
• /
• pp.652-654
• /
• 2023

The accurate prediction of User Equipment (UE) paths in wireless networks is crucial for improving handover mechanisms and optimizing network performance, particularly in the context of Beyond 5G and 6G networks. This paper presents a comprehensive evaluation of unidirectional and bidirectional recurrent neural network (RNN) architectures for UE path prediction. The study employs a sequence-to-sequence model designed to forecast user paths in a wireless network environment, comparing the performance of unidirectional and bidirectional RNNs. Through extensive experimentation, the paper highlights the strengths and weaknesses of each RNN architecture in terms of prediction accuracy and computational efficiency. These insights contribute to the development of more effective predictive path-based mobility management strategies, capable of addressing the challenges posed by ultra-dense cell deployments and complex network dynamics.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.609-614
• /
• 2023

최근 인공지능 시스템이 발전함에 따라 사람보다 높은 성능을 보이고 있다. 또한 전문 지식에 특화된 분야(질병 진단, 법률, 교육 등)에도 적용되고 있지만 이러한 전문 지식 분야는 정확한 판단이 중요하다. 이로 인해 인공지능 모델의 결정에 대한 근거나 해석의 중요성이 대두되었다. 이를 위해 설명 가능한 인공지능 연구인 XAI가 발전하게 되었다. 이에 착안해 본 논문에서는 기계독해 프레임워크에 순환 신경망 디코더를 활용하여 정답 뿐만 아니라 예측에 대한 근거를 추출하고자 한다. 실험 결과, 모델의 예측 답변이 근거 문장 내 등장하는지에 대한 실험과 분석을 수행하였다. 이를 통해 모델이 추론 과정에서 예측 근거 문장을 기반으로 정답을 추론한다는 것을 확인할 수 있었다.