• 한국음향학회지
• /
• 제40권4호
• /
• pp.337-346
• /
• 2021

일반적으로 비정상 신호 탐지 연구에서는 데이터 불균형으로 인해 정상 신호 특징을 주된 정보로 사용한다. 본 논문에서는 비정상 신호의 특징을 학습하는 병렬 오토인코더를 이용한 효율적인 비정상 신호 탐지기법을 제안한다. 제안된 동일한 구조로 이루어진 병렬 오토인코더는 정상 신호와 비정상 신호에 대한 특징을 학습하는 정상 복원기와 비정상 복원기로 구성되며, 정상 및 비정상 데이터를 순차적으로 학습함으로써 불균형 데이터 문제를 효율적으로 해결할 수 있다. 뿐만 아니라 보다 높은 탐지성능 향상을 위해서 부가적인 이진 분류기가 추가될 수 있다. 공개된 음향데이터를 이용한 실험결과, 제안된 병렬 탐지모델의 학습시간이 단일 오토인코더 탐지모델과 비교하여 약 1.31 ~ 1.61배 늘어나지만, 최소 22 % 이상의 Area Under Curve(AUC) 향상을 보였다. 또한, 사전에 훈련된 병렬 오토인코더를 이용하여 수중 음향데이터를 전이학습한 결과 수중 비정상 신호 AUC 탐지성능을 93 % 이상 향상시킬 수 있음을 확인하였다.

• ETRI Journal
• /
• 제37권2호
• /
• pp.251-261
• /
• 2015

We present a novel method for real-time automatic license plate detection in high-resolution videos. Although there have been extensive studies of license plate detection since the 1970s, the suggested approaches resulting from such studies have difficulties in processing high-resolution imagery in real-time. Herein, we propose a novel cascade structure, the fastest classifier available, by rejecting false positives most efficiently. Furthermore, we train the classifier using the core patterns of various types of license plates, improving both the computation load and the accuracy of license plate detection. To show its superiority, our approach is compared with other state-of-the-art approaches. In addition, we collected 20,000 images including license plates from real traffic scenes for comprehensive experiments. The results show that our proposed approach significantly reduces the computational load in comparison to the other state-of-the-art approaches, with comparable performance accuracy.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권4호
• /
• pp.2180-2197
• /
• 2019

With the proliferation of the Android malicious applications, malware becomes more capable of hiding or confusing its malicious intent through the use of code obfuscation, which has significantly weaken the effectiveness of the conventional defense mechanisms. Therefore, in order to effectively detect unknown malicious applications on the Android platform, we propose DroidVecDeep, an Android malware detection method using deep learning technique. First, we extract various features and rank them using Mean Decrease Impurity. Second, we transform the features into compact vectors based on word2vec. Finally, we train the classifier based on deep learning model. A comprehensive experimental study on a real sample collection was performed to compare various malware detection approaches. Experimental results demonstrate that the proposed method outperforms other Android malware detection techniques.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2024년도 춘계학술발표대회
• /
• pp.294-297
• /
• 2024

Malware detection has become increasingly critical with the proliferation of end devices. To improve detection rates and efficiency, the research focus in malware detection has shifted towards leveraging machine learning and deep learning approaches. This shift is particularly relevant in the context of the widespread adoption of end devices, including smartphones, Internet of Things devices, and personal computers. Machine learning techniques are employed to train models on extensive datasets and evaluate various features, while deep learning algorithms have been extensively utilized to achieve these objectives. In this research, we introduce TabNet, a novel architecture designed for deep learning with tabular data, specifically tailored for enhancing malware detection techniques. Furthermore, the Synthetic Minority Over-Sampling Technique is utilized in this work to counteract the challenges posed by imbalanced datasets in machine learning. SMOTE efficiently balances class distributions, thereby improving model performance and classification accuracy. Our study demonstrates that SMOTE can effectively neutralize class imbalance bias, resulting in more dependable and precise machine learning models.

• 한국철도학회:학술대회논문집
• /
• 한국철도학회 2000년도 추계학술대회 논문집
• /
• pp.476-483
• /
• 2000

Like other vehicles, the suspension elements of railway rolling stock have influence on running stability and ride quality. Thus, faults detection for suspension elements is important to prevent an accidents of train and to ensure safety against derailment. This study was started to grasp the feasibility of diagnostic method for the suspension elements of bogie without disassembling. Through several tests by running test rig, we found that fault detection for suspension elements was possible. Here, we describe some results.

• 한국철도학회:학술대회논문집
• /
• 한국철도학회 2010년도 춘계학술대회 논문집
• /
• pp.564-570
• /
• 2010

Last train of the vehicle for eht energy saving and improved performance PWM converters ares widely used. In the case of PWM converters by the zero detection system performance depends on whether it can be argued. Zero voltage detectio method of the hardware and software approach is to in this paper, the zero detection methods for hardware and software problems that have occured as a complemnetary technique was expained.

• 전기학회논문지
• /
• 제57권12호
• /
• pp.2220-2227
• /
• 2008

The safety equipments of railway level crossing which are installed at intersections between roads and railway lines prevent level crossing accidents by informing all of the vehicles and pedestrians of approaching trains. The intelligent safety system for level crossing which employs information and communication technology has been developed in USA and Japan, etc. But, in Korea, the relevant research has not been performed. In this paper, we analyze the cause of railway level crossing accidents and the inherent problem of the existing safety equipments. Based on analyzed results, we design the intelligent safety system which prevent collision between a train and a vehicle. This system displays train approaching information in real-time at roadside warning devices, informs approaching train of the detected obstacle in crossing areas, and is interconnected with traffic signal to empty the crossing area before train comes. Especially, we present the video based obstacle detection algorithm and verify its performance with prototype H/W since the abrupt obstacles in crossing areas are the main cause of level crossing accidents. We identify that the presented scheme detects both pedestrian and vehicle with good performance.

• Journal of Information Processing Systems
• /
• 제19권1호
• /
• pp.109-117
• /
• 2023

Defect detection is one of the key factors in fabric quality control. To improve the speed and accuracy of denim fabric defect detection, this paper proposes a defect detection algorithm based on cascading feature extraction architecture. Firstly, this paper extracts these weight parameters of the pre-trained VGG16 model on the large dataset ImageNet and uses its portability to train the defect detection classifier and the defect recognition classifier respectively. Secondly, retraining and adjusting partial weight parameters of the convolution layer were retrained and adjusted from of these two training models on the high-definition fabric defect dataset. The last step is merging these two models to get the defect detection algorithm based on cascading architecture. Then there are two comparative experiments between this improved defect detection algorithm and other feature extraction methods, such as VGG16, ResNet-50, and Xception. The results of experiments show that the defect detection accuracy of this defect detection algorithm can reach 94.3% and the speed is also increased by 1-3 percentage points.

• 한국철도학회:학술대회논문집
• /
• 한국철도학회 2006년도 춘계학술대회 논문집
• /
• pp.756-761
• /
• 2006

• 한국컴퓨터정보학회논문지
• /
• 제25권3호
• /
• pp.19-26
• /
• 2020

일반적으로 품질 관리는 많은 제조 공정, 특히 주조 또는 용접과 관련된 공정의 기본 구성 요소가 된다. 그러나 사람이 일일이 수동으로 품질 관리 절차를 하는 것은 종종 시간이 걸리고 오류가 발생하기 쉽다. 최근 고품질 제품에 대한 요구를 만족시키기 위해 지능형 육안 검사 시스템의 사용이 생산 라인에서 필수적이 되고 있다. 본 논문에서는 이를 위해 딥 러닝 기반의 ShuffleDefectNet 결함 감지 시스템을 제안하고자 한다. 제안된 결함 검출 시스템은 NEU 데이터 세트의 결함 검출에 대한 여러 최신 성능들보다 높은 평균 정확도 99.75% 정도를 얻는다. 이 논문에서 여러 다른 트레이닝 데이터로부터 최상의 성능을 탐지하고 탐지 성능을 관찰하였다. 그 결과 ShuffleDefectNet의 전체 아키텍처를 사용할 때 정확성과 속도가 크게 향상됨을 알 수 있었다.