• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2021.06a
• /
• pp.105-108
• /
• 2021

Generative adversarial networks (GANs) have reached a great result at creating the synthesis image, especially in the face generation task. Unlike other deep learning tasks, the input of GANs is usually the random vector sampled by a probability distribution, which leads to unstable training and unpredictable output. One way to solve those problems is to employ the label condition in both the generator and discriminator. CelebA and FFHQ are the two most famous datasets for face image generation. While CelebA contains attribute annotations for more than 200,000 images, FFHQ does not have attribute annotations. Thus, in this work, we introduce a method to learn the attributes from CelebA then predict both soft and hard labels for FFHQ. The evaluated result from our model achieves 0.7611 points of the metric is the area under the receiver operating characteristic curve.

• Journal of KIISE:Information Networking
• /
• v.30 no.2
• /
• pp.270-281
• /
• 2003

ETF made the RFCs of MPLS technologies for providing the QoS of ATM or Frame Relay and the flexibility&scalability of IP on the Internet services. IETF has been expanding MPLS technologies as a common control component for supporting the various switching technologies called GMPLS. Also, IETF has standardized the signaling protocols based on such technologies, such as LDP, CR-LDP and RSVP-TE. ETRI developed the MPLS system based on ATM switch in order to provide more reliable services, differentiated services and value-added services like the VPN and traffic engineering service on the Korea Public Sector network. We are planning on deploying model services and commercial services on that network. This paper explains the basic functions of LDP, design and development of LDP on our system, and compares with LDP development and operation on other MPLS systems made by Cisco, Juniper, Nortel and Riverstone. In conclusion, this paper deduces the future services and applications by LDP through these explanation and comparison.

• The KIPS Transactions:PartC
• /
• v.9C no.1
• /
• pp.65-72
• /
• 2002

In this paper, we implement a high-speed swatch tilth the function for label integration to enhance the expansion of networks using the label space of routers efficiently on MPLS over ATM networks. We propose an appropriate hardware structure to support the VC-merge function and differentiated services simultaneously. In this paper, we use the adaptive congestion control method such as EPD algorithm in carte that there is a possibility of network congestion in output buffers of each core LSR. In addition, we justify the validity of the proposed VC-merge method through simulation and comparison to conventional Non VC-merge methods. The proposed VC-merge capable switch is modeled in VHDL. synthesized, and fabricated using the SAMSUNG 0.5um SOG process.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.12
• /
• pp.77-84
• /
• 2022

In this paper, we propose a Label Reassignment Strategy to improve the performance of an object detection algorithm. Our approach involves two stages: an inference stage and an assignment stage. In the inference stage, we perform multi-scale inference with predefined scale sizes on a trained model and re-infer masked images to obtain robust classification results. In the assignment stage, we calculate the IoU between bounding boxes to remove duplicates. We also check box and class occurrence between the detection result and annotation label to re-assign the dominant class type. We trained the YOLOX-L model with the re-annotated dataset to validate our strategy. The model achieved a 3.9% improvement in mAP and 3x better performance on AP_S compared to the model trained with the original dataset. Our results demonstrate that the proposed Label Reassignment Strategy can effectively improve the performance of an object detection model.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.661-664
• /
• 2018

온라인 게시판 글과 채팅창에서 주고받는 대화는 실제 사용되고 있는 구어체 특성이 잘 반영된 텍스트 코퍼스로 음성인식의 언어 모델 재료로 활용하기 좋은 학습 데이터이다. 하지만 온라인 특성상 노이즈가 많이 포함되어 있기 때문에 학습에 직접 활용하기가 어렵다. 본 논문에서는 사용자 입력오류가 다수 포함된 문장에서의 한글 오류 보정을 위한 sequence-to-sequence Denoising Autoencoder 모델을 제안한다.

• Convergence Security Journal
• /
• v.12 no.1
• /
• pp.35-42
• /
• 2012

Increase of data traffic and the advent of new real-time services require to change from the traditional TDM-based (Time Division Multiplexing) networks to the optical networks that soft and dynamic configuration. Voice and lease line services are main service area of the traditional TDM-based networks. This optical network became main infrastructure that offer many channel that can convey data, video, and voice. To provide high resilience against failures, Packet-optical networks must have an ability to maintain an acceptable level of service during network failures. Fast and resource optimized lightpath restoration strategies are urgent requirements for the near future Packet-optical networks with a Generalized Multi-Protocol Label Switching(GMPLS) control plane. The goal of this paper is to provide packet-optical network with a hierarchical multi-layer recovery in order to fast and coordinated restoration in packet-optical network/GMPLS, focusing on new implementation information. The proposed schemes do not need an extension of optical network signaling (routing) protocols for support.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.19 no.6
• /
• pp.227-233
• /
• 2019

Recently, along with the recent development of deep learning technique, neural networks are achieving success in computer vision filed. Convolutional neural network have shown outstanding performance in not only for a simple image classification task, but also for tasks with high difficulty such as object segmentation and detection. However many such deep learning models are based on supervised-learning, which requires more annotation labels than image-level label. Especially image semantic segmentation model requires pixel-level annotations for training, which is very. To solve these problems, this paper proposes a weakly-supervised semantic segmentation method which requires only image level label to train network. Existing weakly-supervised learning methods have limitations in detecting only specific area of object. In this paper, on the other hand, we use multi-classifier deep learning architecture so that our model recognizes more different parts of objects. The proposed method is evaluated using VOC 2012 validation dataset.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.7C
• /
• pp.677-684
• /
• 2002

Today's Internet does not have efficient traffic engineering mechanism to support QoS for the explosive increasing internet traffic such as various multimedia traffic. This functional shortage degrades prominently the quality of service, and makes it difficult to provide multi-media service and real-time service. Various technologies are under developed to solve these problems. IETF (Internet Engineering Task Force) developed the MPLS (Multi-Protocol Label Switching) technology that provides a good capabilities of traffic engineering and is independent layer 2 protocol, so MPLS is expected to be used in the Internet backbone network$\^$[1][2]/. The faults occurring in high-speed network such as MPLS, may cause massive data loss and degrade quality of service. So fast network restoration function is essential requirement. Because MPLS is independent to layer 2 protocol, the fault detection and reporting mechanism for restoration should also be independent to layer 2 protocol. In this paper, we present the experimental results of the MPLS OAM function for the performance monitoring and fault detection 'll'&'ll' notification, localization in MPLS network, based on the OPNET network simulator

• The Journal of Bigdata
• /
• v.7 no.2
• /
• pp.1-9
• /
• 2022

In the global era, the importance of imported food safety management is increasing. Address information of overseas food companies is key information for imported food safety management, and must be verified for prompt response and follow-up management in the event of a food risk. However, because each country's address system is different, one verification system cannot verify the addresses of all countries. Also, the purpose of address verification may be different depending on the field used. In this paper, we deal with the problem of classifying a given overseas food business address into the administrative district level of the country. This is because, in the event of harm to imported food, it is necessary to find the administrative district level from the address of the relevant company, and based on this trace the food distribution route or take measures to ban imports. However, in some countries the administrative district level name is omitted from the address, and the same place name is used repeatedly in several administrative district levels, so it is not easy to accurately classify the administrative district level from the address. In this study we propose a deep learning-based administrative district level classification model suitable for this case, and verify the actual address data of overseas food companies. Specifically, a method of training using a label powerset in a multi-label classification model is used. To verify the proposed method, the accuracy was verified for the addresses of overseas manufacturing companies in Ecuador and Vietnam registered with the Ministry of Food and Drug Safety, and the accuracy was improved by 28.1% and 13%, respectively, compared to the existing classification model.

• Annual Conference on Human and Language Technology
• /
• 2017.10a
• /
• pp.56-59
• /
• 2017

CNN(Convolutional Neural Network)을 이용하여 발화 주제 다중 분류 task를 multi-labeling 방법과, cluster 방법을 이용하여 수행하고, 각 방법론에 MSE(Mean Square Error), softmax cross-entropy, sigmoid cross-entropy를 적용하여 성능을 평가하였다. Network는 음절 단위로 tokenize하고, 품사정보를 각 token의 추가한 sequence와, Naver DB를 통하여 얻은 named entity 정보를 입력으로 사용한다. 실험결과 cluster 방법으로 문제를 변형하고, sigmoid를 output layer의 activation function으로 사용하고 cross entropy cost function을 이용하여 network를 학습시켰을 때 F1 0.9873으로 가장 좋은 성능을 보였다.