• 로봇학회논문지
• /
• 제14권2호
• /
• pp.122-130
• /
• 2019

Collecting a rich but meaningful training data plays a key role in machine learning and deep learning researches for a self-driving vehicle. This paper introduces a detailed overview of existing open-source simulators which could be used for training self-driving vehicles. After reviewing the simulators, we propose a new effective approach to make a synthetic autonomous vehicle simulation platform suitable for learning and training artificial intelligence algorithms. Specially, we develop a synthetic simulator with various realistic situations and weather conditions which make the autonomous shuttle to learn more realistic situations and handle some unexpected events. The virtual environment is the mimics of the activity of a genuine shuttle vehicle on a physical world. Instead of doing the whole experiment of training in the real physical world, scenarios in 3D virtual worlds are made to calculate the parameters and training the model. From the simulator, the user can obtain data for the various situation and utilize it for the training purpose. Flexible options are available to choose sensors, monitor the output and implement any autonomous driving algorithm. Finally, we verify the effectiveness of the developed simulator by implementing an end-to-end CNN algorithm for training a self-driving shuttle.

• 한국실천공학교육학회논문지
• /
• 제2권1호
• /
• pp.120-125
• /
• 2010

우선선정직종훈련제도는 우리나라 전체 직업훈련 재정투자의 10% 내외를 차지하고 있는 직업능력개발사업의 핵심사업 중의 하나이다. 우선선정직종훈련제도는 2010년 5월 31일의 근로자직업능력개발법의 개정에 따라 "국가 기간 전략산업직종"의 훈련제도로 그 성격은 명확해졌으나 여전히 우선선정직종훈련제도의 훈련직종의 선정과정이 산업수요를 제대로 반영하기 어려운 문제점이 있으며, 민간위탁훈련기관의 선정 및 관리방식에 많은 문제점을 안 고 있다. 이에 본 논문에서는 현행 우선선정직종훈련제도가 가지고 있는 이러한 문제점을 해소하여 국가기간 전략산업직종의 핵심인력을 양성하는 제도로서 발전해나가기 위한 방향을 모색하였다. 구체적으로는 국가기간 전략산업직종에 대한 훈련수요 파악 및 직종 선정을 지역의 훈련기관과 산업계를 중심으로 이루어지도록 사회적 인프라를 구축하며, 민간위탁훈련기관의 선정, 계약, 관리 등을 개선해나갈 것을 제안하였다.

• 한국잠사학회:학술대회논문집
• /
• 한국잠사학회 2004년도 International Conference of Sericultural and Entomological Sciences
• /
• pp.28-29
• /
• 2004

• 대한음성학회지:말소리
• /
• 제43호
• /
• pp.45-55
• /
• 2002

High reliability of automatic speaker recognition regardless of voice conditions is necessary for forensic application. Audio recordings in real cases are not consistent in voice conditions, such as duration, time interval of recording, given text or conversational speech, transmission channel, etc. In this study the variation of verification error rate of ASR system with the voice conditions was investigated. As a result in order to decrease both false rejection rate and false acception rate, the various voices should be used for training and the duration of train voices should be longer than the test voices.

• 한국항해학회지
• /
• 제14권4호
• /
• pp.53-70
• /
• 1990

Much efforts have been made to improve the training education system for last decades. however, it still leaves much room form improving the system. The reason for this is that the have been many changes in given educational conditions, national and international, and that there existed the lack of training facilities on shore and the limits of capacity on the training ship. The existing program adopts a straight-through system of which the course has to be completed at same time, and also forces students to study the course, disregarding their aptitude for sea life. Consequently, the program resulted in frustrating the learning desire of some students and, as a consequence, in deteriorating the quality of the entire training education. This paper aims to develop an efficient training program including curriculla by the literature survey and the teaching and sea experiences on the training ship "HANBADA" and merchant ships, where the authors have been for many years. Compared with the existing one, the new training model suggested in this paper has some advantages as follows : First, the new model adopts multi-state system which consists of various short-term training courses according to each purpose. This system will be helpful for student to find their aptitude for sea life earlier and to understand classes of major subjection shore. Second, the model includes new curriculla which consist of core subjects (for example, navigation, marine operation, marine transportation, watch keeping and nautical English for deck cadets and internal and external combustion engine, auxiliary machinery, electric and electronics and engine maintenance for engine cadets), by incorporating existing 20 subjects in 5 subjects. These curriculla may contribute to embodying the characteristics of training education where the above mentioned subjects must be linked with each other. In order to implement this new training model efficiently and effectively, the following prerequisties must be prepared : $\circled1$ The contents of each subject included in the new model should be systematically developed. $\circled2$ The educational schedule should be adjusted according to the new model.new model.

• 대한한의진단학회지
• /
• 제11권2호
• /
• pp.13-26
• /
• 2007

Background: Neurofeedback is an electroencephalographic biofeedback technique for training individuals to alter their brain activity via operant conditioning. Also neurofeedback is a form of behavioural training aimed at developing skills for brain activity. Within the past decade, several neurofeedback studies have been published that tend to overcome the methodological shortcomings of earlier studies. This research describes the methodical basis of neurofeedback and reviews the evidence base for its clinical efficacy and effectiveness in attention-deficit hyperactivity disorder (ADHD). Methods: In neurofeedback training, self-regulation of specific aspects of electrical brain activity is acquired by means of immediate feedback and positive reinforcement. In frequency training, activity in different EEG frequency bands has to be decreased or increased. Slow cortical potentials (SCPs) training is focused on the regulation of cortical excitability. Results: Neurofeedback studies revealed training-specific effects on, for example, attention and memory processes and performance improvements in real-life conditions, in healthy subjects as well as in patients. In several studies it was shown that ADHD symptomatology was reduced after frequency training or SCP(Slow cortical potentials) training. Moreover a decrease of impulsivity errors and an increase of the contingent negative variation. Conclusions: This research provides evidence for both positive behavioural and specific neurophysiological effects of neurofeedback training. Also there is growing evidence for neurofeedback as a valuable module in neuropsychiatric disorders. Further, controlled studies are warranted.

• 한국컴퓨터정보학회논문지
• /
• 제27권1호
• /
• pp.107-113
• /
• 2022

본 논문에서는 홈트레이닝의 기대감과 운동몰입, 운동지속의도와의 관계에 대해 제안한다. 연구의 대상은 2021년 3월 현재 부산, 경남 지방경찰청 소속 경찰공무원을 대상으로 한다. 그 중 홈트레이닝에 참여해 본 경험이 있는 337명을 대상으로 하여 설문조사를 실시하였다. 연구결과, 첫째 홈트레이닝의 기대감은 운동몰입, 운동지속의도 요인과 유의한 상관이 있는 것으로 나타났다. 둘째, 홈트레이닝 참여자의 기대감은 운동몰입, 운동지속의도에 유의한 영향을 미치는 것으로 나타났다. 셋째, 홈트레이닝 참여자의 운동몰입은 운동지속의도에 유의한 영향을 미치는 것으로 조사되었다. 각종 치안현장에 대비한 체력조건을 준비하기 위해 홈트레이닝은 유용하게 활용될 것이다. 단기간의 트레이닝이 아닌 장기간의 계획을 세워 트레이닝을 실시한다면 홈트레이닝이 추구하는 효과를 충분히 달성할 수 있을 것이다.

• 한국운동역학회지
• /
• 제33권1호
• /
• pp.17-24
• /
• 2023

Objective: The purpose of this study was to investigate the effects of different frequency on of knee extensors muscle function during electrical muscle stimulation (EMS). Method: In this research, 40 subjects who have no musculoskeletal disorder, and less than a year workout experience were recruited in order to analyze effects of EMS with different stimulus frequency. Forty subjects were randomly divided into four groups of ten subjects in each group. A EMS training program with different frequencies (without EMS [WE], EMS with frequency 30 Hz [E30], EMS with frequency 60 Hz [E60], EMS with frequency 90 Hz [E90]) was assigned to each group. Throughout eight weeks of training, test subjects were simultaneously carried out knee extension exercises such as squat, leg extension, and leg-press while using EMS with different frequency (20 min, pulse width 250 ㎲, on-off ratio 1:1). Isokinetic knee extension strength, muscle activity of the rectus femoris (RF), the vastus medialis (VM), and the vastus lateralis (VL), and the median frequency of the RF, the VM, and the VL were collected and compared between pre and post training in order to find effects of applying EMS with different frequencies. For each dependent variable, a one-way ANOVA was to determine whether there were significant differences among four different conditions (p<.05). When a significant difference was found, post hoc analyses were performed using the contrast procedure. Results: When compared to WE and E90, E30 causes significant increase in isokinetic knee extension strength. No significant differences were found in EMG values across different EMS conditions. However, the median frequency of the VM in E30 was significantly increased than the corresponding value for WE. Conclusion: The results of this study showed that EMS training with 30 Hz frequency had positive effect on knee extensor. Based of the findings of the present study, EMS training with lower frequency may help the performer to focus on developing strength in knee extensor muscles.

• 한국정밀공학회지
• /
• 제20권11호
• /
• pp.64-70
• /
• 2003

This paper suggests a systematic methodology to predict chip forms using the experimental design technique and the neural network. Significant factors determined with ANOVA analysis are used as input variables of the neural network back-propagation algorithm. It has been shown that cutting conditions and cutting tool shapes have distinct effects on the chip forms, so chip breaking. Cutting tools are represented using the Z-map method, which differs from existing methods using some chip breaker parameters. After training the neural network with selected input variables, chip forms are predicted and compared with original chip forms obtained from experiments under same input conditions, showing that chip forms are same at all conditions. To verify the suggested model, one tool not used in training the model is chosen and input to the model. Under various cutting conditions, predicted chip forms agree well with those obtained from cutting experiments. The suggested method could reduce the cost and time significantly in designing cutting tools as well as replacing the“trial-and-error”design method.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권10호
• /
• pp.4930-4951
• /
• 2018

The deep learning based speech enhancement has shown considerable success. However, it still suffers performance degradation under mismatch conditions. In this paper, an adaptation method is proposed to improve the performance under noise mismatch conditions. Firstly, we advise a noise aware training by supplying identity vectors (i-vectors) as parallel input features to adapt deep neural network (DNN) acoustic models with the target noise. Secondly, given a small amount of adaptation data, the noise-dependent DNN is obtained by using $L_2$ regularization from a noise-independent DNN, and forcing the estimated masks to be close to the unadapted condition. Finally, experiments were carried out on different noise and SNR conditions, and the proposed method has achieved significantly 0.1%-9.6% benefits of STOI, and provided consistent improvement in PESQ and segSNR against the baseline systems.