This paper proposes a data augmentation algorithm to improve the performance of DNN(Deep Neural Network) based speech enhancement. Many deep learning models are exploring algorithms to maximize the performance in limited amount of data. The most commonly used algorithm is the data augmentation which is the technique artificially increases the amount of data. For the effective data augmentation algorithm, we used a formant enhancement method that assign the different weights to the formant frequencies. The DNN model which is trained using the proposed data augmentation algorithm was evaluated in various noise environments. The speech enhancement performance of the DNN model with the proposed data augmentation algorithm was compared with the algorithms which are the DNN model with the conventional data augmentation and without the data augmentation. As a result, the proposed data augmentation algorithm showed the higher speech enhancement performance than the other algorithms.

Since a surgery simulation should be able to represent the internal structure of the human body, it is advantageous to adopt volume based techniques rather than polygon based techniques. However, the volume based techniques induce large computation to deform heterogeneous volume datasets such as bones and muscles. In this study, we propose a new method to deform volume data using multi-core CPUs. By improving previous studies, the proposed method minimizes unnecessary propagation operations. Moreover, we propose an efficient task-partitioning method for volume deformation using multi-core CPUs. As a result, we can simulate the deformation of heterogeneous volume data at an interactive speed without special hardware.

Video coding technologies are progressively becoming more efficient and complex. The Versatile Video Coding (VVC) is a new state-of-the art video compression standard that is going to be a standard, as the next generation of High Efficiency Video Coding (HEVC) standard. To explore the future video coding technologies beyond the HEVC, numerous efficient methods have been adopted by the Joint Video Exploration Team (JVET). Since then, the next generation video coding standard named as VVC and its software model called VVC Test Model (VTM) have emerged. In this paper, several important coding features for motion estimation and motion compensation in the VVC standard is introduced and analyzed in terms of the performance. Improved coding tools introduced for ME and MC in VVC, can achieve much better and good balance between coding efficiency and coding complexity compared with the HEVC.

In this paper, we propose a new indoor localization method for indoor mobile robots using LiDAR. The indoor mobile robots operating in limited areas usually require high-precision localization to provide high level services. The performance of the widely used localization methods based on radio waves or computer vision are highly dependent on their usage environment. Therefore, the reproducibility of the localization is insufficient to provide high level services. To overcome this problem, we propose a new localization method based on the comparison between ceiling shape information obtained from LiDAR measurement and the blueprint. Specifically, the method includes a reliable segmentation method to classify point clouds into connected planes, an effective comparison method to estimate position by matching 3D point clouds and 2D blueprint information. Since the ceiling shape information is rarely changed, the proposed localization method is robust to its usage environment. Simulation results prove that the position error of the proposed localization method is less than 10 cm.

The effects of different heat treatments on the sensory characteristics of abalones are studied in this study. In this paper, the sensory evaluation of abalone samples under different heat treatment conditions is carried out, and the evaluation results are analyzed. The three-dimensional (3D) scanning and reverse engineering are used in tooth modeling of the sensory evaluation of abalone samples under different heat treatment conditions. Besides, the chewing movement models are simplified into three modes, including the cutting mode, compressing mode and grinding mode, which are simulated using finite element simulation. The elastic modulus of the abalone samples is obtained through the compression testing using a texture analyzer to distinguish their material properties under different heat treatments and to obtain simulated mechanical parameters. Finally, taking the mechanical parameters of the finite element simulation of abalone chewing as input and sensory evaluation parameters as the output, BP neural network is established in which the sensory texture evaluation model of abalone samples is obtained. Through verification, the neural network prediction model can meet the requirements of food texture evaluation, with an average error of 9.12%.

The patient's waiting time and length of stay have been reported as a factor decreasing their satisfaction in the hospital, especially in developing countries. This paper focuses on modeling hospital's outpatient department workflow in a developing country and optimizing the patient waiting time as well as total length of stay. By using discrete-event simulation, many alternative scenarios have raised, such as adding more working time, altering human resources, and adjusting the staff's responsibility, those scenarios will be examined to explore better settings for the hospital. The results show that possible to achieve a 9.6% reduction in patient total length of stay and it could be accomplished without adding more resources to the hospital.

This paper presents the design and implementation of a Mission Planning System(MPS) for multiple ballistic missiles. MPSs are also a kind of M&S systems in defense domain, and these provide important computations on the ground before flights of flying objects. The proposed MPS in this paper has a new concept which is far from generating a set of waypoints of a flying object and proving the set. In this paper, we firstly discuss the research motivation of our own MPS; then, we introduce the design of our MPS and its functionalities. In order to prove the practicality of our MPS, we have conducted a case study.