• Annual Conference of KIPS
• /
• 2023.05a
• /
• pp.81-83
• /
• 2023

태양 에너지 수집형 IoT 기기는 주기적으로 재충전되는 태양 에너지의 특성상, 에너지 소모를 최소화하기보다는 수집된 에너지를 최대한 유용하게 사용하는 것이 중요하다. 한편, 데이터 기밀성과 프라이버시, 응답속도, 비용 등의 이유로 클라우드가 아닌 데이터 소스 근처에서 머신러닝을 수행하는 엣지 AI에 대한 연구도 활발한데, 그 중 하나는 여러 IoT 장치들이 수집한 오디오 데이터를 활용하여, 다양한 AI 응용들을 IoT 엣지 컴퓨팅 환경에서 제공하는 것이다. 그러나, 이와 관련된 많은 연구에서, IoT 기기들은 에너지의 제약으로 인하여, 엣지 서버(IoT 서버)로의 센싱 데이터 전송만을 수행하고, 데이터 전처리를 포함한 모든 AI 과정은 엣지 서버에서 수행한다. 이 경우, 엣지 서버의 과부하 문제 뿐 아니라, 학습 및 추론에 불필요한 데이터까지도 서버에 그대로 전송되므로 네트워크 과부하 문제도 야기한다. 또한, 이를 해결하고자, 데이터 전처리 과정을 각 IoT 기기에 모두 맡긴다면, 기기의 에너지 부족으로 정전시간이 증가하는 또 다른 문제가 발생한다. 본 논문에서는 각 IoT 기기의 에너지 상태에 따라 데이터 전처리 여부를 결정함으로써, 기기들의 정전시간 증가 문제를 완화시키면서 서버 집중형 엣지 AI 환경의 문제들(엣지 서버 및 네트워크 과부하)을 완화시키고자 한다. 제안기법에서 IoT 장치는 기기가 기본적으로 동작하는 데 필요한 에너지 외의 여분의 에너지 양을 예측하고, 이 여분의 에너지가 있는 경우에만 이를 사용하여 기기에서 전처리 과정, 즉 수집 대상 소리 판별과 잡음 제거 과정을 거친 후 서버에 전송함으로써, IoT기기의 정전시간에 영향을 주지 않으면서, 에너지 적응적으로 데이터 전처리 위치(IoT기기 또는 엣지 서버)를 결정하여 수행한다.

• Economic and Environmental Geology
• /
• v.55 no.3
• /
• pp.309-316
• /
• 2022

In order to minimize the human and time consumption required for rock classification, research on rock classification using artificial intelligence (AI) has recently developed. In this study, basic volcanic rocks were subdivided by using polarizing microscope thin section images. A convolutional neural network (CNN) model based on Tensorflow and Keras libraries was self-producted for rock classification. A total of 720 images of olivine basalt, basaltic andesite, olivine tholeiite, trachytic olivine basalt reference specimens were mounted with open nicol, cross nicol, and adding gypsum plates, and trained at the training : test = 7 : 3 ratio. As a result of machine learning, the classification accuracy was over 80-90%. When we confirmed the classification accuracy of each AI model, it is expected that the rock classification method of this model will not be much different from the rock classification process of a geologist. Furthermore, if not only this model but also models that subdivide more diverse rock types are produced and integrated, the AI model that satisfies both the speed of data classification and the accessibility of non-experts can be developed, thereby providing a new framework for basic petrology research.

• Journal of Intelligence and Information Systems
• /
• v.29 no.3
• /
• pp.37-55
• /
• 2023

With the development of electronic journals and the emergence of various interdisciplinary studies, the selection of journals for publication has become a new challenge for researchers. Even if a paper is of high quality, it may face rejection due to a mismatch between the paper's topic and the scope of the journal. While research on assisting researchers in journal selection has been actively conducted in English, the same cannot be said for Korean journals. In this study, we propose a system that recommends Korean journals for submission. Firstly, we utilize SBERT (Sentence BERT) to embed abstracts of previously published papers at the document level, compare the similarity between new documents and published papers, and recommend journals accordingly. Next, the order of recommended journals is determined by considering the similarity of abstracts, keywords, and title. Subsequently, journals that are similar to the top recommended journal from previous stage are added by using a dictionary of words constructed for each journal, thereby enhancing recommendation diversity. The recommendation system, built using this approach, achieved a Top-10 accuracy level of 76.6%, and the validity of the recommendation results was confirmed through user feedback. Furthermore, it was found that each step of the proposed framework contributes to improving recommendation accuracy. This study provides a new approach to recommending academic journals in the Korean language, which has not been actively studied before, and it has also practical implications as the proposed framework can be easily applied to services.

• Journal of Internet Computing and Services
• /
• v.19 no.6
• /
• pp.41-51
• /
• 2018

Recently, in the field of video surveillance, deep learning based learning method is applied to intelligent video surveillance system, and various events such as crime, fire, and abnormal phenomenon can be robustly detected. However, since occlusion occurs due to the loss of 3d information generated by projecting the 3d real-world in 2d image, it is need to consider the occlusion problem in order to accurately detect the object and to estimate the pose. Therefore, in this paper, we detect moving objects by solving the occlusion problem of object detection process by adding depth information to existing RGB information. Then, using the convolution neural network in the detected region, the positions of the 14 keypoints of the human joint region can be predicted. Finally, in order to solve the self-occlusion problem occurring in the pose estimation process, the method for 3d human pose estimation is described by extending the range of estimation to the 3d space using the predicted result of 2d keypoint and the deep neural network. In the future, the result of 2d and 3d pose estimation of this research can be used as easy data for future human behavior recognition and contribute to the development of industrial technology.

• Journal of Navigation and Port Research
• /
• v.47 no.6
• /
• pp.358-366
• /
• 2023

This study aimed to provide a solution for improving ship collision alert of the 'accident vulnerable ship monitoring service' among the 'intelligent marine traffic information system' services of the Ministry of Oceans and Fisheries. The current ship collision alert uses a supervised learning (SL) model with survey labels based on large ship-oriented data and its operators. Consequently, the small ship data and the operator's opinion are not reflected in the current collision-supervised learning model, and the effect is insufficient because the alarm is provided from a longer distance than the small ship operator feels. In addition, the supervised learning (SL) method requires a large number of labeled data, and the labeling process requires a lot of resources and time. To overcome these limitations, in this paper, the classification model of collision alerts for small ships using unlabeled data with the semi-supervised learning (SSL) algorithms (Label Propagation and TabNet) was studied. Results of real-time experiments on small ship operators using the classification model of collision alerts showed that the satisfaction of operators increased.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.17 no.5
• /
• pp.1-8
• /
• 2017

Recently, with the development of ICT technology such as cloud and mobile, image utilization through social networks is increasing rapidly. These images contain personal information, and personal information leakage accidents may occur. As a result, studies are underway to recognize and mask personal information in images. However, optical character recognition, which recognizes personal information in images, varies greatly depending on brightness, contrast, and distortion, and Korean recognition is insufficient. Therefore, in this paper, we design and implement a personal information identification and masking system based on image recognition through deep learning application using CNN algorithm based on optical character recognition method. Also, the proposed system and optical character recognition compares and evaluates the recognition rate of personal information on the same image and measures the face recognition rate of the proposed system. Test results show that the recognition rate of personal information in the proposed system is 32.7% higher than that of optical character recognition and the face recognition rate is 86.6%.

• Journal of Broadcast Engineering
• /
• v.25 no.1
• /
• pp.13-24
• /
• 2020

There are about nine million power line poles and 1.3 million kilometers of the power line for electric power distribution in Korea. Maintenance of such a large number of electric power facilities requires a lot of manpower and time. Recently, various fault diagnosis techniques using artificial intelligence have been studied. Therefore, in this paper, proposes a power line deflection detect system using artificial intelligence and computer vision technology in images taken by vision system. The proposed system proceeds as follows. (i) Detection of transmission tower using object detection system (ii) Histogram equalization technique to solve the degradation in image quality problem of video data (iii) In general, since the distance between two transmission towers is long, a panoramic video stitching process is performed to grasp the entire power line (iv) Detecting deflection using computer vision technology after applying power line detection algorithm This paper explain and experiment about each process.

• Journal of Broadcast Engineering
• /
• v.25 no.5
• /
• pp.776-788
• /
• 2020

In this paper, a vehicle type recognition system using deep learning and a license plate recognition system are proposed. In the existing system, the number plate area extraction through image processing and the character recognition method using DNN were used. These systems have the problem of declining recognition rates as the environment changes. Therefore, the proposed system used the one-stage object detection method YOLO v3, focusing on real-time detection and decreasing accuracy due to environmental changes, enabling real-time vehicle type and license plate character recognition with one RGB camera. Training data consists of actual data for vehicle type recognition and license plate area detection, and synthetic data for license plate character recognition. The accuracy of each module was 96.39% for detection of car model, 99.94% for detection of license plates, and 79.06% for recognition of license plates. In addition, accuracy was measured using YOLO v3 tiny, a lightweight network of YOLO v3.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.8
• /
• pp.1248-1254
• /
• 2022

Efforts are being made to prevent traffic accidents in the school zone in advance. However, traffic accidents in school zones continue to occur. If the driver can know the situation information in the child protection area in advance, accidents can be reduced. In this paper, we design a camera that eliminates blind spots in school zones and a number recognition camera system that can collect pre-traffic information. It is designed by improving the LIDAR system that recognizes vehicle speed and pedestrians. It collects and processes pedestrian and vehicle image information recognized by cameras and LIDAR, and applies artificial intelligence time series analysis and artificial intelligence algorithms. The artificial intelligence traffic accident prevention system learned by deep learning proposed in this paper provides a forced push service that delivers school zone information to the driver to the mobile device in the vehicle before entering the school zone. In addition, school zone traffic information is provided as an alarm on the LED signboard.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.101-105
• /
• 2018

이 논문은 가장 잘 알려진 어휘부중 하나인 워드넷의 활용 범위 확장을 위해 워드넷 신셋에 "사건구조 프레임(Event Structure Frame)"을 주석하는 연구에 관한 것이다. 워드넷을 비롯하여 현재 사용되고 있는 어휘부는 풍부한 어휘의미정보가 구조화되어 있지만, 사건구조에 관한 정보를 포함하고 있지는 않다. 이 연구의 가장 큰 기여는 워드넷에 사건구조 프레임을 추가함으로써 워드넷과의 연결만으로 핵심적인 어휘의미정보를 모두 추출할 수 있도록 해준다는 점이다. 예를 들어 텍스트 추론, 자연어처리, 멀티 모달 태스크 등은 어휘의미정보와 배경지식(상식)을 이용하여 태스크를 수행한다. 워드넷에 대한 사건구조 주석은 자동사건구조 주석 시스템인 GESL을 이용하여 워드넷 신셋에 있는 예문에 먼저 자동 주석을 하고, 오류에 대해 수동 수정을 하는 반자동 방식이다. 사전 정의된 23개의 사건구조 프레임에 따라 예문에 출현하는 타겟 동사를 분류하고, 해당 프레임과 매핑한다. 현재 이 연구는 시작 단계이며, 이 논문에서는 빈도 순위가 가장 높은 100개의 동사와 각 사건구조 프레임별 대표 동사를 포함하여 총 106개의 동사 레마에 대해 실험을 진행하였다. 그 동사들에 대한 전체 워드넷 신셋의 수는 1337개이다. 예문이 없어서 GESL이 적용될 수 없는 신셋을 제외하면 1112개 신셋이다. 이 신셋들에 대해 GESL을 적용한 결과 F-Measure는 73.5%이다. 향후 연구에서는 워드넷-사건구조 링크를 계속 업데이트하면서 딥러닝을 이용해 GESL 성능을 향상 할 수 있는 방법을 모색할 것이다.