• Journal of information and communication convergence engineering
• /
• v.8 no.6
• /
• pp.640-644
• /
• 2010

When we talk with hands-free in a car or noisy lobby, the performance of the echo canceller degrade because background noise added to echo caused by the distance from mouth to microphone is relatively long. It gives a reason for necessity of noise-robust and high convergence speed adaptive algorithm. And if acoustic echo canceller operated not perfectly, residual signal going through the echo canceller to far-end speaker remains residual echo, which degrade quality of talk. To solve this problem, post-processing needed to remove residual echo ones more. In this paper, we propose a new acoustic echo canceller, which has noise robust and high convergence speed, linked with linear predictor as a post-processor. By computer simulation, it is confirmed that the proposed algorithm shows better performance from acoustic interference cancellation (AIC) viewpoint.

• Journal of Information Processing Systems
• /
• v.17 no.2
• /
• pp.385-398
• /
• 2021

The complexity of deep learning models affects the real-time performance of gesture recognition, thereby limiting the application of gesture recognition algorithms in actual scenarios. Hence, a residual learning neural network based on a deep convolutional neural network is proposed. First, small convolution kernels are used to extract the local details of gesture images. Subsequently, a shallow residual structure is built to share weights, thereby avoiding gradient disappearance or gradient explosion as the network layer deepens; consequently, the difficulty of model optimisation is simplified. Additional convolutional neural networks are used to accelerate the refinement of deep abstract features based on the spatial importance of the gesture feature distribution. Finally, a fully connected cascade softmax classifier is used to complete the gesture recognition. Compared with the dense connection multiplexing feature information network, the proposed algorithm is optimised in feature multiplexing to avoid performance fluctuations caused by feature redundancy. Experimental results from the ISOGD gesture dataset and Gesture dataset prove that the proposed algorithm affords a fast convergence speed and high accuracy.

• The Journal of Korea Robotics Society
• /
• v.16 no.1
• /
• pp.72-78
• /
• 2021

This paper presents a robust impedance control of high-DOF robot based on disturbance observer(DOB). A novel DOB is derived by considering the residual disturbance caused by the difference between actual disturbance and disturbance decoupling input which utilizes the estimated disturbance. It focuses on the elimination of the residual disturbance and improvement of the control performance as well as the good estimation of disturbances. In the control of high-DOF robot, numerical dynamic model, which is conducted by a software based on dynamics, is utilized because the analytical model of high-DOF robot is difficult to be obtained. The simulation of high-DOF robot with numerical dynamic model is provided to verify the performance of the proposed controller.

• Journal of Information Processing Systems
• /
• v.18 no.1
• /
• pp.12-25
• /
• 2022

In recent years, deep convolutional neural networks have made significant progress in the research of single image super-resolution. However, it is difficult to be applied in practical computing terminals or embedded devices due to a large number of parameters and computational effort. To balance these problems, we propose CSRNet, a lightweight neural network based on channel split residual learning structure, to reconstruct highresolution images from low-resolution images. Lightweight refers to designing a neural network with fewer parameters and a simplified structure for lower memory consumption and faster inference speed. At the same time, it is ensured that the performance of recovering high-resolution images is not degraded. In CSRNet, we reduce the parameters and computation by channel split residual learning. Simultaneously, we propose a double-upsampling network structure to improve the performance of the lightweight super-resolution network and make it easy to train. Finally, we propose a new evaluation metric for the lightweight approaches named 100_FPS. Experiments show that our proposed CSRNet not only speeds up the inference of the neural network and reduces memory consumption, but also performs well on single image super-resolution.

• Journal of Advanced Information Technology and Convergence
• /
• v.10 no.1
• /
• pp.111-120
• /
• 2020

Generative adversarial networks (GANs) achieved impressive performance on image generation and visual classification applications. However, adversarial networks meet difficulties in combining the generative model and unstable training process. To overcome the problem, we combined the deep residual network with upsampling convolutional layers to construct the generative network. Moreover, the study shows that image generation and classification performance become more prominent when the residual layers include on the generator. The proposed network empirically shows that the ability to generate images with higher visual accuracy provided certain amounts of additional complexity using proper regularization techniques. Experimental evaluation shows that the proposed method is superior to image generation and classification tasks.

• Journal of Information Processing Systems
• /
• v.19 no.3
• /
• pp.323-333
• /
• 2023

Facial expression recognition can aid in the development of fatigue driving detection, teaching quality evaluation, and other fields. In this study, a facial expression recognition method was proposed with a residual masking reconstruction network as its backbone to achieve more efficient expression recognition and classification. The residual layer was used to acquire and capture the information features of the input image, and the masking layer was used for the weight coefficients corresponding to different information features to achieve accurate and effective image analysis for images of different sizes. To further improve the performance of expression analysis, the loss function of the model is optimized from two aspects, feature dimension and data dimension, to enhance the accurate mapping relationship between facial features and emotional labels. The simulation results show that the ROC of the proposed method was maintained above 0.9995, which can accurately distinguish different expressions. The precision was 75.98%, indicating excellent performance of the facial expression recognition model.

• Journal of Welding and Joining
• /
• v.33 no.1
• /
• pp.30-34
• /
• 2015

Residual stresses are inherently introduced into the engineering components during manufacturing including rolling, forging, bending and welding processes. Excessive residual stresses are known to be detrimental to the proper integrity and performance of components. Neutron diffraction has become a well-established technique for the determination of residual stresses in welds. The deep penetration capability of neutrons into most metallic materials makes neutron diffraction a powerful tool for the residual stress measurements through the thickness of the weld specimen. Furthermore, the unique volume-averaged bulk characteristic of the scattering beam and mapping capability in three dimensions is suitable for the engineering purpose. In this presentation, the neutron diffraction measurements of the residual stresses will be introduced and measurement results will highlighted in thick weld plates.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.10 no.6
• /
• pp.11-18
• /
• 2002

Residual gas fraction in an engine cylinder affects engine performance, efficiency and emission characteristics. With high residual gas fractions, a flame speed and maximum combustion temperature are decreased and these are deeply related with combustion stability especially at idle and NOx emission at relatively high engine load. In this work, the residual gas fraction was calculated by an engine simulation code, which was validated by the experimental data (cylinder pressure and emissions) obtained from 4-cyliner spark ignition engine. A comparison between experimental and computational calculation results was made. The residual gas is generated mostly at low engine speed by the larger pressure difference between the intake and exhaust port. As the valve overlap duration was increased, the amount of residual gas in the cylinder, the amount of HC emission in the exhaust gas and the variation of power output increased.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.17 no.10
• /
• pp.2437-2445
• /
• 1993

In this work, a new method is presented for generating the manipulator path which significantly reduces residual vibration under the torque constraints. The desired path is optimally designed so that the required movement can be achieved with minimum residual vibration. From the previous research works, the dynamic model had been established including both the link and the joint flexibilities. The performance index is selected to minimize the maximum amplitude of residual vibration. The path to be designed is developed by a combined Fourier series and polynomial function to satisfy both the convergence and boundary condition matching problems. The concept of correlation coefficients is used to select the minimum number of design variables, i.e. Fourier coefficients, the only ones which have a considerable effect on the reduction of residual vibration. A two-link Manipulator is used to evaluate this method. Results show that residual vibration can be drastically reduced by selecting an appropriate manipulator path to both of unlimited and torque-limited cases.

• Earthquakes and Structures
• /
• v.25 no.1
• /
• pp.15-26
• /
• 2023

It is of great importance to assess the residual displacement demand in the performance-based seismic design. In this paper, a hybrid deep learning model for predicting the residual displacement spectra under near-fault (NF) ground motions is proposed by combining the long short-term memory network (LSTM) and back-propagation (BP) network. The model is featured by its capacity of predicting the residual displacement spectrum under a given NF ground motion while considering the effects of structural parameters. To construct this model, 315 natural and artificial NF ground motions were employed to compute the residual displacement spectra through elastoplastic time history analysis considering different structural parameters. Based on the resulted dataset with a total of 9,450 samples, the proposed model was finally trained and tested. The results show that the proposed model has a satisfactory accuracy as well as a high efficiency in predicting residual displacement spectra under given NF ground motions while considering the impacts of structural parameters.