• Journal of Internet Computing and Services
• /
• v.20 no.4
• /
• pp.91-102
• /
• 2019

In autonomous driving systems, the ability to classify pedestrians in images captured by cameras is very important for pedestrian safety. In the past, after extracting features of pedestrians with HOG(Histogram of Oriented Gradients) or SIFT(Scale-Invariant Feature Transform), people classified them using SVM(Support Vector Machine). However, extracting pedestrian characteristics in such a handcrafted manner has many limitations. Therefore, this paper proposes a method to classify pedestrians reliably and effectively using CNN's(Convolutional Neural Network) deep features and transfer learning. We have experimented with both the fixed feature extractor and the fine-tuning methods, which are two representative transfer learning techniques. Particularly, in the fine-tuning method, we have added a new scheme, called M-Fine(Modified Fine-tuning), which divideslayers into transferred parts and non-transferred parts in three different sizes, and adjusts weights only for layers belonging to non-transferred parts. Experiments on INRIA Person data set with five CNN models(VGGNet, DenseNet, Inception V3, Xception, and MobileNet) showed that CNN's deep features perform better than handcrafted features such as HOG and SIFT, and that the accuracy of Xception (threshold = 0.5) isthe highest at 99.61%. MobileNet, which achieved similar performance to Xception and learned 80% fewer parameters, was the best in terms of efficiency. Among the three transfer learning schemes tested above, the performance of the fine-tuning method was the best. The performance of the M-Fine method was comparable to or slightly lower than that of the fine-tuningmethod, but higher than that of the fixed feature extractor method.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.20 no.5
• /
• pp.972-978
• /
• 2016

MACsec is a cryptographic function that operates on Layer 2, the international standard defined in the IEEE 802.1AE. As industries such as IoT(Internet of Things) devices are receiving attention recently are connected to the network and Internet traffic is increasing rapidly, and is exposed to the risk of a variety of Internet attacks. Traditional network security technologies were often made in Layer 3, such as IPsec. However, to be increased as rapidly as the current traffic situation is complicated, and became interested in the security function of protecting the entire traffic instead of for a specific application or protocol. It appeared as these technologies is technology MACsec technology to protect all traffic in Layer 2. In this paper, we propose a Layer 2 security technology adapter MACsec MACsec a technology that allows you to simply and easily add them to the existing Layer 2 networks.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.39C no.5
• /
• pp.435-446
• /
• 2014

Even if it is important to transmit high rate multimedia information without any transmission errors for surveillance systems, it is difficult to achieve error-free transmission due to infra-less adhoc networks. In order to reduce the transmission errors furthermore, additional signal overheads or retransmission of signals should be required, but they may lead to transmission delay. This paper represents a study on low delay and adaptive video transmission for the unmanned surveillance systems by developing system protocols. In addition, we introduce an efficient and adaptive control algorithm using system parameters for exploiting unmanned surveillance system properly over multi-channels.

• The KIPS Transactions:PartC
• /
• v.11C no.7 s.96
• /
• pp.905-916
• /
• 2004

Self-organizing hierarchical ad hoc network (SOHAN) is a new ad-hoc network architecture designed to improve the scalability properties of conventional 'flat' ad hoc networks. This network architecture consists of three tiers of ad-hoc nodes, i.e.. access points, forwarding nodes and mobile nodes. This paper presents a topology discovery and routing protocol for the self-organization of SOHAN. We propose a cross-layer path metric based on link quality and MAC delay which plays a key role in producing an optimal cluster-based hierarchical topology with high throughput capacity. The topology discovery protocol provides the basis for routing which takes place in layer 2.5 using MAC addresses. The routing protocol is based on AODV with appropriate modifications to take advantage of the hierarchical topology and interact with the discovery protocol. Simulation results are presented which show the improved performance as well as scalability properties of SOHAN in terms of through-put capacity, end-to-end delay, packet delivery ratio and control overhead.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2011.08a
• /
• pp.301-301
• /
• 2011

최근 고밀도 메모리 반도체의 재료와 빠른 응답을 요구하는 나노입자를 이용한 비휘발성 메모리 소자의 제작에 대한 연구가 활발히 진행되고 있다. 그에 따른 기존의 플래쉬 메모리가 가지는 문제점을 개선하기 위해서 균일하고 규칙적으로 분포하는 새로운 나노소재의 개발과 비휘발성, 고속 동작, 고집적도, 저전력 소자의 공정기술이 요구되고 있다. 또한 부유게이트에 축적되는 저장되는 전하량을 증가시키기 위한 새로운 소자구조 개발이 필요하다. 한편, 실리 사이드 계열의 나노입자는 금속 나노입자와 달리 현 실리콘 기반의 반도체 공정에서 장점을 가지고 있다. 따라서 본 연구에서는 화합물 중에서 비휘발성 메모리 장치의 전기적 특성을 향상 시킬 수 있는 실리사이드 계열의 바나듐 실리사이드(V3Si) 박막을 열처리 과정을 통하여 수 nm 크기의 나노입자로 제작하였다. 소자의 제작은 p-Si기판에 실리콘산화막 터널층(5 nm 두께)을 건식 산화법으로 성장 후, 바나듐 실리사이드 금속박막을 RF 마그네트론 스퍼터 시스템을 이용하여 4~6 nm 두께로 터널 베리어 위에 증착하고, 그 위에 초고진공 마그네트론 스퍼터링을 이용하여 SiO2 컨트롤 산화막층 (20 nm)을 형성시켰다. 여기서 V3Si 나노입자 형성을 위해 급속 열처리법으로 질소 분위기에서 800$^{\circ}C$로 5초 동안 열처리하여 하였으며, 마지막으로 열 기화 시스템을 통하여 알루미늄 전극(직경 200 ${\mu}m$, 두께 200 nm)을 증착하여 소자를 제작하였다. 제작된 구조는 금속 산화막 반도체구조를 가지는 나노 부유게이트 커패시터이며, 제작된 시편은 투사전자현미경을 이용하여 나노입자의 크기와 균일성을 확인했다. 소자의 전기적인 측정을 E4980A capacitor parameter analyzer와 Agilent 81104A pulse pattern generator system을 이용한 전기용량-전압 측정을 통해 전하저장 효과 및 메모리 동작 특성들을 분석하고, 열처리 조건에 따라 형성되는 V3Si 의 조성을 엑스선 광전자 분광법을 이용하여 확인하였다.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.19 no.5
• /
• pp.178-188
• /
• 2020

This paper proposes a lane recognition CNN network using split-attention network as a backbone to extract feature. Split-attention is a method of assigning weight to each channel of a feature map in the CNN feature extraction process; it can reliably extract the features of an image during the rapidly changing driving environment of a vehicle. The proposed deep neural networks in this paper were trained and evaluated using the Tusimple data set. The change in performance according to the number of layers of the backbone network was compared and analyzed. A result comparable to the latest research was obtained with an accuracy of up to 96.26, and FN showed the best result. Therefore, even in the driving environment of an actual vehicle, stable lane recognition is possible without misrecognition using the model proposed in this study.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.42 no.6
• /
• pp.63-72
• /
• 2005

In this paper, we propose the bit-rate control algorithm for the block based image compression like H.263, H.263+ or MPEG-4. The proposed algorithm is designed to identify the quantization parameter set through the Lagrangian optimization technique based on the well-known linear source model. We set the Lagrangian cost function with the rates and distortion calculated from the linear source model. We calculate the quantization parameter set using the Vitervi algorithm to solve the Lagrangian optimization problem considering the Dquant method of H.263 and MPEG-4. The proposed algorithm improves the video quality by up to 1.5 dB compared with the TMN8 scheme, and is more effective in the video sources with dynamic activities than the consistent quality approaches.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.5
• /
• pp.906-912
• /
• 2012

A ship ad-hoc network (SANET) can provide ships with diverse multimedia services by replacing expensive satellite communications. While ITU-R M. 1842-1, standards for maritime VHF band digital communications, can be used as the specifications of physical layer for SANETs, no standards are specified for higher layers of SANETs. In this paper, we propose a ship ad-hoc communication (SAC) protocol for SANETs, based on medium access control (MAC) and routing protocols for terrestrial ad-hoc networks. SAC protocol is a cross-layer protocol which combines MAC and routing into one algorithm and considers maritime environments, including the existence of neighboring ships, the possibility of routing to a destination, and changing the communication mode in case of VHF channel failure.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.32 no.5
• /
• pp.817-826
• /
• 2022

AI technology is being successfully introduced in many fields, and models deployed as a service are deployed with black box environment that does not expose the model's information to protect intellectual property rights and data. In a black box environment, attackers try to steal data or parameters used during training by using model output. This paper proposes a method of inferring the type of model to directly find out the composition of layer of the target model, based on the fact that there is no attack to infer the information about the type of model from the deep learning model. With ResNet, VGGNet, AlexNet, and simple convolutional neural network models trained with MNIST datasets, we show that the types of models can be inferred using the output values in the gray box and black box environments of the each model. In addition, we inferred the type of model with approximately 83% accuracy in the black box environment if we train the big and small relationship feature that proposed in this paper together, the results show that the model type can be infrerred even in situations where only partial information is given to attackers, not raw probability vectors.

• Journal of Advanced Navigation Technology
• /
• v.27 no.5
• /
• pp.495-501
• /
• 2023

P-3C/K aircraft for maritime patrols that Republic of Korea Navy is using, is equipped with a variety of sensors and communication devices. Collected data from the aircraft is managed as tactical information by flight operators and stored. When the flight mission is completed, this information is transferred to tactical support center on the ground and played back or used for follow-up work through a analysis tool. During a flight mission, there are tens of thousands of detection objects within an hour in KADIZ (Korea air defense identification zone). In contrast, in TSD (tactical situation display), which displays a map when using the analysis tool, all detected objects are expressed as symbols. The increase in display symbols has a significant impact on the TSD image updating and consequently interferes with the smooth operation of operators. In this paper, we propose applying multiple threads and multiple layers to improve the performance of existing TSD. And the performance improvement is proven through the execution results.