• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.1
• /
• pp.1-15
• /
• 2022

The MobileNetV3 is specially designed for mobile devices with limited memory and computing power. To reduce the network parameters and improve the network inference speed, a new lightweight network is proposed based on MobileNetV3. Firstly, to reduce the computation of residual blocks, a partial residual structure is designed by dividing the input feature maps into two parts. The designed partial residual structure is used to replace the residual block in MobileNetV3. Secondly, a dual-path feature extraction structure is designed to further reduce the computation of MobileNetV3. Different convolution kernel sizes are used in the two paths to extract feature maps with different sizes. Besides, a transition layer is also designed for fusing features to reduce the influence of the new structure on accuracy. The CIFAR-100 dataset and Image Net dataset are used to test the performance of the proposed partial residual structure. The ResNet based on the proposed partial residual structure has smaller parameters and FLOPs than the original ResNet. The performance of improved MobileNetV3 is tested on CIFAR-10, CIFAR-100 and ImageNet image classification task dataset. Comparing MobileNetV3, GhostNet and MobileNetV2, the improved MobileNetV3 has smaller parameters and FLOPs. Besides, the improved MobileNetV3 is also tested on CPU and Raspberry Pi. It is faster than other networks

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.21 no.5
• /
• pp.27-32
• /
• 2021

This paper evaluates the equalization performance of FC-MMA adaptive equalization algorithm by the fixed step size that is used for the minimization of the intersymbol interference which occurs in the time dispersive communication channel. The FC-MMA has a fast convergence speed in order to adapts the new environment more rapidly in case of the time varying charateristics and the abnormal situation like as outage of the communication channel. But the algorithms operates in adative method, convegence speed is depend on fixed step size for adaptation. For this situation, its performance was evaluated by changing the step size value, the residual isi and maximum distortion and MSE performance index which means the convergence characteristics are widely adapted in the adaptive equalizer, SER were applied. As a result of computer simulation, the large step size can improves the convergence speed for reaching the steady state, but has a poor performance compared to small step size in residual values after steady state. The research result shows that the FC-MMA algorithm is applied the large step size for rapidly reaching the steady state in initial time, then adjust the small step size after reaching the steady state for reducing the residual values for equalization.

• Journal of Advanced Navigation Technology
• /
• v.18 no.5
• /
• pp.455-460
• /
• 2014

The system performance is analized for the optimal design of the transmission links with dispersion management and optical phase conjugation for compensating for the optical signal distortion due to the group velocity dispersion and optical nonlinear Kerr effects in the long-haul optical transmission system. That is, the effect of the relation of the residual dispersion in both half transmission sections with respect with optical phase conjugator (OPC) on the net residual dispersion (NRD) is assessed. It is conformed that the best compensation is obtained in NRD of 10 ps/nm, which is only controlled by the difference of the residual dispersion between each half transmission sections.

• Nuclear Engineering and Technology
• /
• v.49 no.4
• /
• pp.817-824
• /
• 2017

Residual stress is a critical element in determining the integrity of parts and the lifetime of welded structures. It is necessary to estimate the residual stress of a welding zone because residual stress is a major reason for the generation of primary water stress corrosion cracking in nuclear power plants. That is, it is necessary to estimate the distribution of the residual stress in welding of dissimilar metals under manifold welding conditions. In this study, a cascaded support vector regression (CSVR) model was presented to estimate the residual stress of a welding zone. The CSVR model was serially and consecutively structured in terms of SVR modules. Using numerical data obtained from finite element analysis by a subtractive clustering method, learning data that explained the characteristic behavior of the residual stress of a welding zone were selected to optimize the proposed model. The results suggest that the CSVR model yielded a better estimation performance when compared with a classic SVR model.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.15 no.2
• /
• pp.99-104
• /
• 2015

This paper proposes the RMMA (Region based Multiple Modulus Algorithm) algorithm that is possible to improving the performance of MMA adaptive equalization algorithm in order to the reduction of intersymbol interference occurs at the communication channel. In RMMA algorithm, the output constellation of equalizer are divided by 4 different regions in order to get the error signal for adapting the channel characteristic, and the small error signal is obtained by mapping each region to 4-QAM signal. The conversion effect of constant modulus from nonconstant modulus signal was obtained. In this paper, the adaptive equalization performance of proposed RMMA were evaluated comared to the present MMA. As a result of computer simulation, the convergence speed and residual quantity were improved in residual isi and MD. Especially the superiorities of robustness was confirm in SER performance compared to present MMA.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.23 no.5
• /
• pp.1-8
• /
• 2009

Solenoid valve cleaner is used for clean the solenoid valve in automatic transmission for a car when it is clogged with transmission oil sludge. Nevertheless when the solenoid is turned off, the residual current in the solenoid coil makes slow motion of the plunger in the solenoid which makes lower in cleansing performance and speed. In this paper, the method of performance Improvement for solenoid valve cleaner is proposed. The residual current in the solenoid coil is eliminated rapidly by the proposed method and it improves the cleansing performance and speed. The experimental results show the validity of the reposed method.

• Journal of the Society of Naval Architects of Korea
• /
• v.59 no.4
• /
• pp.243-250
• /
• 2022

In the design stage of hull forms, a fast prediction method of resistance performance is needed. In these days, large test matrix of candidate hull forms is tested using Computational Fluid Dynamics (CFD) in order to choose the best hull form before the model test. This process requires large computing times and resources. If there is a fast and reliable prediction method for hull form performance, it can be used as the first filter before applying CFD. In this paper, we suggest the offset-based performance prediction method. The hull form geometry information is applied in the form of 2D offset (non-dimensionalized by breadth and draft), and it is studied using Convolutional Neural Network (CNN) and adapted to the model test results (Residual Resistance Coefficient; C_R). Some additional variables which are not included in the offset data such as main dimensions are merged with the offset data in the process. The present model shows better performance comparing with the simple regression models.

• Proceedings of the KIEE Conference
• /
• 1998.07d
• /
• pp.1519-1521
• /
• 1998

A Gapless HST(high speed train) arrester design is not possible without the highly non-linear ZnO(ZincOxide) varistors. Zno varistors combine excellent protective characteristics with steady state performance to maximize protection, the ZnO varistors are selected for each unit based on leakage current and residual voltage, to verify that the residual voltage is the residual voltage published for HST arrester.

• Proceedings of the Korea Society for Energy Engineering kosee Conference
• /
• 1997.10a
• /
• pp.9-16
• /
• 1997

원유 정제의 가장 heavy한 잔류물인 중잔유(heavy residual oil)의 IGCC 프랜트의 적용성능을 평가하기 위한 방안으로, 정적시스템 모사방법을 사용하여 중잔유를 발전 연료로 사용한 IGCC 플랜트를 모사하였다. 모사에 적용한 중잔유는 Visbreaker Residue와Butane Asphalt이며, 시스템 모사방법의 검증을 위해서, 중잔유의 가스화 반응 모사결과를 Shell사에서 발표한 실증자료와 비교하여 사용된 모사방법이 적절함을 입증한 후 이 결과를 이용하여 IGCC 플랜트에 대한 모사에 적용하였다.

• Journal of Broadcast Engineering
• /
• v.24 no.4
• /
• pp.633-642
• /
• 2019

Recently, the convolution neural network (CNN) model at a single image super-resolution (SISR) have been very successful. The residual learning method can improve training stability and network performance in CNN. In this paper, we propose a SISR using recursive residual network architecture by introducing dense skip connections for learning nonlinear mapping from low-resolution input image to high-resolution target image. The proposed SISR method adopts a method of the recursive residual learning to mitigate the difficulty of the deep network training and remove unnecessary modules for easier to optimize in CNN layers because of the concise and compact recursive network via dense skip connection method. The proposed method not only alleviates the vanishing-gradient problem of a very deep network, but also get the outstanding performance with low complexity of neural network, which allows the neural network to perform training, thereby exhibiting improved performance of SISR method.