• Convergence Security Journal
• /
• v.21 no.2
• /
• pp.57-66
• /
• 2021

Deep neural networks (DNNs) provide excellent performance for image, speech, and pattern recognition. However, DNNs sometimes misrecognize certain adversarial examples. An adversarial example is a sample that adds optimized noise to the original data, which makes the DNN erroneously misclassified, although there is nothing wrong with the human eye. Therefore studies on defense against adversarial example attacks are required. In this paper, we have experimentally analyzed the success rate of detection for adversarial examples by adjusting various parameters. The performance of the ensemble defense method was analyzed using fast gradient sign method, DeepFool method, Carlini & Wanger method, which are adversarial example attack methods. Moreover, we used MNIST as experimental data and Tensorflow as a machine learning library. As an experimental method, we carried out performance analysis based on three adversarial example attack methods, threshold, number of models, and random noise. As a result, when there were 7 models and a threshold of 1, the detection rate for adversarial example is 98.3%, and the accuracy of 99.2% of the original sample is maintained.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.20 no.2
• /
• pp.95-111
• /
• 2021

As the population decreases in an aging society, the average age of drivers increases. Accordingly, the elderly at high risk of being in an accident need autonomous-driving vehicles. In order to secure driving safety on the road, several technologies to respond to various obstacles are required in those vehicles. Among them, technology is required to recognize static obstacles, such as poor road conditions, as well as dynamic obstacles, such as vehicles, bicycles, and people, that may be encountered while driving. In this study, we propose a deep neural network algorithm capable of simultaneously detecting these two types of obstacle. For this algorithm, we used 1,418 road images and produced annotation data that marks seven categories of dynamic obstacles and labels images to indicate road damage. As a result of training, dynamic obstacles were detected with an average accuracy of 46.22%, and road surface damage was detected with a mean intersection over union of 74.71%. In addition, the average elapsed time required to process a single image is 89ms, and this algorithm is suitable for personal mobility vehicles that are slower than ordinary vehicles. In the future, it is expected that driving safety with personal mobility vehicles will be improved by utilizing technology that detects road obstacles.

• Journal of the Korea Computer Graphics Society
• /
• v.25 no.3
• /
• pp.11-20
• /
• 2019

In this study, a novel virtual reality (VR) experience environment is proposed for enabling walking adaptation of visually impaired people. The core of proposed VR environment is based on immersive walking interactions and deep learning based braille blocks recognition. To provide a realistic walking experience from the perspective of visually impaired people, a tracker-based walking process is designed for determining the walking state by detecting marching in place, and a controller-based VR white cane is developed that serves as the walking assistance tool for visually impaired people. Additionally, a learning model is developed for conducting comprehensive decision-making by recognizing and responding to braille blocks situated on roads that are followed during the course of directions provided by the VR white cane. Based on the same, a VR application comprising an outdoor urban environment is designed for analyzing the VR walking environment experience. An experimental survey and performance analysis were also conducted for the participants. Obtained results corroborate that the proposed VR walking environment provides a presence of high-level walking experience from the perspective of visually impaired people. Furthermore, the results verify that the proposed learning algorithm and process can recognize braille blocks situated on sidewalks and roadways with high accuracy.

• Journal of Digital Convergence
• /
• v.17 no.11
• /
• pp.207-213
• /
• 2019

The fuel-cut coast-down driving mode is activated when the acceleration pedal is released with transmission gear engaged, and it's a default function for electronic-controlled engine of vehicles. The fuel economy becomes better because fuel injection stops during fuel-cut driving mode. A fuel-cut detection method is suggested in the study and it's based on the speed, acceleration and road gradient data from GPS sensor. It detects fuel-cut driving mode by comparing calculated acceleration and realtime acceleration value. The one is estimated with driving resistance in the condition of fuel-cut driving and the other is from GPS sensor. The detection accuracy is about 80% when the method is verified with road driving data. The result is estimated with 9,600 data set of vehicle speed, acceleration, fuel consumption and road gradient from test driving on the road of 12km during 16 minutes, and the road slope is rather high. It's easy to detect fuel-cut without injector signal obtained by connecting wire. The detection error is from the fact that the variation range of speed, acceleration and road gradient data, used for road resistance force, is larger than the value of fuel consumption data.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.12
• /
• pp.63-71
• /
• 2020

In this paper, we propose a fire detection technology using YOLOv3 and EfficientDet, the most reliable artificial intelligence detection algorithm recently, an alert service that simultaneously transmits four kinds of notifications: text, web, app and e-mail, and an AWS system that links fire detection and notification service. There are two types of our highly accurate fire detection algorithms; the fire detection model based on YOLOv3, which operates locally, used more than 2000 fire data and learned through data augmentation, and the EfficientDet, which operates in the cloud, has conducted transfer learning on the pretrained model. Four types of notification services were established using AWS service and FCM service; in the case of the web, app, and mail, notifications were received immediately after notification transmission, and in the case of the text messaging system through the base station, the delay time was fast enough within one second. We proved the accuracy of our fire detection technology through fire detection experiments using the fire video, and we also measured the time of fire detection and notification service to check detecting time and notification time. Our AI fire detection and notification service system in this paper is expected to be more accurate and faster than past fire detection systems, which will greatly help secure golden time in the event of fire accidents.

• Journal of Korean Society of Forest Science
• /
• v.100 no.3
• /
• pp.374-381
• /
• 2011

The purpose of this study is to examine the possibility of early detection of Pine Wilt Disease (PWD) using NDVI (Normalized Difference Vegetation Index) from ADC (Agricultural Digital Camera) imageries. The PWD induces the different patterns of reduction of NDVI between healthy trees and infected trees, due to the withered leaves on the infected trees. Based on these phenomena, the DI showing the NDVI variations of trees by time series was employed to detect the infected trees. To find out the differences of DI level between normal and infected trees, DIs of trees from May to August in 2007 were calculated and these were analyzed with GLM (General Linear Models) in SAS 9.2. As a result, the difference of DI between in June and August shows the most significant level (0.0001). The discriminant analysis was performed between normal and infected trees, using the DI of June and August. As the result, hit ratio of trees and the accuracy of grouping with Jack-knife method were shown 71.9% and 73.5%, respectively. These results showed that the DI is effective to detect the trees infected by the PWD and it is useful to prevent the PWD.

• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.101-125
• /
• 2022

Recently, with the development of computing technology and the improvement of the cloud environment, deep learning technology has developed, and attempts to apply deep learning to various fields are increasing. A typical example is anomaly detection, which is a technique for identifying values or patterns that deviate from normal data. Among the representative types of anomaly detection, it is very difficult to detect a contextual anomaly that requires understanding of the overall situation. In general, detection of anomalies in image data is performed using a pre-trained model trained on large data. However, since this pre-trained model was created by focusing on object classification of images, there is a limit to be applied to anomaly detection that needs to understand complex situations created by various objects. Therefore, in this study, we newly propose a two-step pre-trained model for detecting abnormal situation. Our methodology performs additional learning from image captioning to understand not only mere objects but also the complicated situation created by them. Specifically, the proposed methodology transfers knowledge of the pre-trained model that has learned object classification with ImageNet data to the image captioning model, and uses the caption that describes the situation represented by the image. Afterwards, the weight obtained by learning the situational characteristics through images and captions is extracted and fine-tuning is performed to generate an anomaly detection model. To evaluate the performance of the proposed methodology, an anomaly detection experiment was performed on 400 situational images and the experimental results showed that the proposed methodology was superior in terms of anomaly detection accuracy and F1-score compared to the existing traditional pre-trained model.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_1
• /
• pp.991-1005
• /
• 2022

Satellite imaging is an effective supplementary data source for detecting and verifying nuclear activity. It is also highly beneficial in regions with limited access and information, such as nuclear installations. Time series analysis, in particular, can identify the process of preparing for the conduction of a nuclear experiment, such as relocating equipment or changing facilities. Differences in the semantic segmentation findings of time series photos were employed in this work to detect changes in meaningful items connected to nuclear activity. Building, road, and small object datasets made of KOMPSAT 3/3A photos given by AIHub were used to train deep learning models such as U-Net, PSPNet, and Attention U-Net. To pick relevant models for targets, many model parameters were adjusted. The final change detection was carried out by including object information into the first change detection, which was obtained as the difference in semantic segmentation findings. The experiment findings demonstrated that the suggested approach could effectively identify altered pixels. Although the suggested approach is dependent on the accuracy of semantic segmentation findings, it is envisaged that as the dataset for the region of interest grows in the future, so will the relevant scope of the proposed method.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_4
• /
• pp.1925-1934
• /
• 2022

Monitoring nuclear activity for inaccessible areas using remote sensing technology is essential for nuclear non-proliferation. In recent years, deep learning has been actively used to detect nuclear-activity-related small objects. However, high-resolution satellite imagery containing small objects can result in class imbalance. As a result, there is a performance degradation problem in detecting small objects. Therefore, this study aims to improve detection accuracy by analyzing the effect of the ratio of small objects related to nuclear activity in the input data for the performance of the deep learning model. To this end, six case datasets with different ratios of small object pixels were generated and a U-Net model was trained for each case. Following that, each trained model was evaluated quantitatively and qualitatively using a test dataset containing various types of small object classes. The results of this study confirm that when the ratio of object pixels in the input image is adjusted, small objects related to nuclear activity can be detected efficiently. This study suggests that the performance of deep learning can be improved by adjusting the object pixel ratio of input data in the training dataset.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.5
• /
• pp.1-9
• /
• 2022

Localization systems can be used with various circumstances like measuring population movement and rescue technology, even in security technology (like infiltration detection system). Vision sensors such as camera often used for localization is susceptible with light and temperature, and can cause invasion of privacy. In this paper, we used ultra-wideband radar technology (which is not limited by aforementioned problems) and machine learning techniques to measure the number and location of occupants in other indoor spaces behind the wall. We used four different algorithms and compared their results, including extremely randomized tree for four different situations; detect the number of occupants in a classroom, split the classroom into 28 locations and check the position of occupant, select one out of the 28 locations, divide it into 16 fine-grained locations, and check the position of occupant, and checking the positions of two occupants (existing in different locations). Overall, four algorithms showed good results and we verified that detecting the number and location of occupants are possible with high accuracy using machine learning. Also we have considered the possibility of service expansion using the oneM2M standard platform and expect to develop more service and products if this technology is used in various fields.