• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 한국정보통신학회 2022년도 춘계학술대회
• /
• pp.399-401
• /
• 2022

A definitive diagnosis to classify the breast malignancy status may be achieved by microscopic analysis using surgical open biopsy. However, this procedure requires experts in the specializing of histopathological image analysis directing to time-consuming and high cost. To overcome these issues, deep learning is considered practically efficient to categorize breast cancer into benign and malignant from histopathological images in order to assist pathologists. This study presents a pre-trained convolutional neural network model architecture with a 100% fine-tuning scheme and Adagrad optimizer to classify the breast cancer histopathological images into benign and malignant using a 40× magnification BreaKHis dataset. The pre-trained architecture was constructed using the InceptionResNetV2 model to generate a modified InceptionResNetV2 by substituting the last layer with dense and dropout layers. The results by demonstrating training loss of 0.25%, training accuracy of 99.96%, validation loss of 3.10%, validation accuracy of 99.41%, test loss of 8.46%, and test accuracy of 98.75% indicated that the modified InceptionResNetV2 model is reliable to predict the breast malignancy type from histopathological images. Future works are necessary to focus on k-fold cross-validation, optimizer, model, hyperparameter optimization, and classification on 100×, 200×, and 400× magnification.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• 제46권8호
• /
• pp.59-66
• /
• 2009

Recently, as the expectation about the IPTV (Internet Protocol TV) service quality is rapidly increased by the development of the national high-speed internet and TPS (the Triple Play Service : data + image + audio) service Therefore, the enactment of the national quality standards about the IPTV service quality guaranteeing the real time video quality of a subscriber and the international standards are hastily needed. This paper built a test bed with the network domain and the subscriber set-top box domain including the headend area and commercial network characteristic in order to test in the environment which is similar to the characteristic of the service business network. And by using the constructed environment, the characteristics required for SLA(service Level Agreement) of the IPTV service are presented as the quality test according to the service environment change.

• Journal of Internet Computing and Services
• /
• 제8권5호
• /
• pp.45-58
• /
• 2007

This paper presents UI4GSD (User Interface for Grid Service Development) for grid service developers that provides an efficient and friendly development environment based on Globus Toolkit 4, Normally, implementing grid service requires special expert knowledge for Grid as well as programming, Moreover, grid service development as well as testing of the deployed service in the Globus container takes a long Time, which makes the grid service implementation very inefficient, However, UI4GSD can automatically generate a grid service interface file, a build file, and grid service class files as well as a client class file, using the information supplied by developers through GUI. In UI4GSD, a grid service is developed easily based on typical five step processes with required input data fed at each step. As, a result, UI4GSD can provide an easy and convenient development environment thereby increasing the efficiency and convenience in developing grid services.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• 제14권10호
• /
• pp.2261-2268
• /
• 2010

The information captured by imaging devices such as CCD or CIS may contain external noises through the processes of passing signals or storing images. In this paper, we propose a Modified Sigma Filter (MSF) algorithm to reduce such noises. In experiment, we verified that our MSF algorithm showed better performance in PSNR and 1D plot of simulation results compared with Gaussian Filter (GF), Local Sigma Filter (LSF). Tested images include random Gaussian Noises.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• 제18권11호
• /
• pp.718-724
• /
• 2017

Considerable efforts have been made to expand the boundaries of domestic offshore plant industries, which have focused on the construction of the structures, to the engineering field. On the other hand, time domain analysis, which is one of the most important areas in designing floating offshore plants, relies mainly on the information given by foreign companies. As an early design of the Dynamic Positioning System (DPS) is mostly conducted by several specialized companies, domestic ship builders need to spend time and money to reflect the analysis into the hull shape design. This paper presents the framework required to analyze time domain performance of floating type offshore structures, which are equipped with DPS. To easily perform time domain analysis, framework generates the required input data for the solver, and is modularized to test the control algorithm and performance of a certain DPS. The effectiveness of the developed framework was verified by a simulation with a model ship and the total time for the entire analysis work was reduced by 50% or more.

• Convergence Security Journal
• /
• 제19권4호
• /
• pp.43-53
• /
• 2019

Recently, a number of military intranet infiltrations suspected of North Korea have been discovered. There was a problem that a vulnerability could occur due to the modification of user authentication data that can access existing military information systems. In this paper, we applied mutual verification technique and API (Application Programming Interface) forgery / forgery blocking and obfuscation to solve the authentication weakness in web browsers that comply with FIDO (Fast IDentity Online) standard. In addition, user convenience is improved by implementing No-Plugin that does not require separate program installation. Performance tests show that most browsers perform about 0.1ms based on the RSA key generation rate. In addition, it proved that it can be used for commercialization by showing performance of less than 0.1 second even in the digital signature verification speed of the server. The service is expected to be useful for improving military information system security as an alternative to browser authentication by building a web secure storage.

• The KIPS Transactions:PartB
• /
• 제13B권5호
• /
• pp.547-552
• /
• 2006

There have been various cohesion measurements studied considering reference relation among attributes and methods in a class. Generally, these cohesion measurement are camed out in one class. If the range of reference relation considered are extended from one class to two classes, we could find out the reference relation between two classes. Tn this paper, we proposed a neural network to determine the method location. Neural network is effective to predict output value from input data not to be included in training and generalize after training input and output pattern repeatedly. Learning vector is generated with 30-dimensional input vector and one target binary values of method location in a constraint that there are two classes which have less than or equal to 5 attributes and methods The result of the proposed neural network is about 95% in cross-validation and 88% in testing.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• 제37권3호
• /
• pp.55-61
• /
• 2000

This paper presents an efficient hardware architecture of frame memory interface in MPEG-2 video encoder. To reduce the size of memory buffers between SDRAM and the frame memory module, the number of clocks needed for each memory access is minimized with dual bank operation and burst length change. By allocating the remaining cycles not used by SDRAM access, to the random access cycle, the internal buffer size, the data bus width, and the size of the control logic can be minimized. The proposed architecture is operated with 54MHz clock and designed with the VT $I^{тм}$ 0.5 ${\mu}{\textrm}{m}$ CMOS TLM standard cell library. It is verified by comparing the test vectors generated by the c-code model with the simulation results of the synthesized circuit. The buffer area of the proposed architecture is reduced to 40 % of the existing architecture.

• Journal of Internet Computing and Services
• /
• 제14권6호
• /
• pp.9-18
• /
• 2013

In this paper, we design and implement an intelligent system for finding objects with RFID(Radio Frequency IDentification) tag in which an mobile robot can do. The system we developed is a learning system of artificial neural network that uses RSSI(Received Signal Strength Indicator) value as input and absolute coordination value as target. Although a passive RFID is used for location estimation, we consider an active RFID for expansion of recognition distance. We design the proposed system and construct the environment for indoor location estimation. The designed system is implemented with software and the result related learning is shown at test bed. We show various experiment results with similar environment of real one from earning data generation to real time location estimation. The accuracy of location estimation is verified by simulating the proposed method with allowable error. We prepare local test bed for indoor experiments and build a mobile robot that can find the objects user want.

• Journal of the Korean Institute of Intelligent Systems
• /
• 제9권5호
• /
• pp.555-565
• /
• 1999

The optimal identification algorithm of fuzzy systems is presented for rule-based fuzzy modeling of nonlinear complex systems. Nonlinear systems are expressed using the identification of structure such as input variables and fuzzy input subspaces, and parameters of a fuzzy model. In this paper, the rule-based fuzzy modeling implements system structure and parameter identification using the fuzzy inference methods and hybrid structure combined with two types of optimization theories for nonlinear systems. Two types of inference methods of a fuzzy model are the simplified inference and linear inference. The proposed hybrid optimal identification algorithm is carried out using both a genetic algorithm and the improved complex method. Here, a genetic algorithm is utilized for determining initial parameters of membership function of premise fuzzy rules, and the improved complex method which is a powerful auto-tuning algorithm is carried out to obtain fine parameters of membership function. Accordingly, in order to optimize fuzzy model, we use the optimal algorithm with a hybrid type for the identification of premise parameters and standard least square method for the identification of consequence parameters of a fuzzy model. Also, an aggregate performance index with weighting factor is proposed to achieve a balance between performance results of fuzzy model produced for the training and testing data. Two numerical examples are used to evaluate the performance of the proposed model.