• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.105-108
• /
• 1999

TTS(Text-To-Speech) 시스템 합성음성의 자연감을 개선하기 위해 하나의 언어에 대해 존재하는 운율 법칙을 정확히 구현해야 한다. 존재하는 운율 법칙을 추출하기 위해서는 방대한 분량의 언어 자료 구축이 필요하다. 그러나 이 방법은 존재하는 운율 현상이 포함된 언어자료에 대해 완벽한 운율을 파악할 수 없으므로 합성음성의 질을 좋게 할 수 없다. 본 논문은 한국어 음성의 운율을 학습하기 위해 2개의 인공 신경망을 제안한다. 하나의 신경망으로 문장의 각 음소에 대한 피치 변화를 학습시키는 것이며, 다른 하나는 에너지 변화를 학습하도록 하였다. 신경망은 BP 신경망을 이용하며 11개의 음소를 나타내기 위해 11개의 입력과, 중간 음소의 피치와 에너지 변화곡선을 근사하는 다항식 계수를 출력하도록 하였다. 신경망시스템의 학습과 평가에 앞서, 음성학적 균형잡힌 고립단어를 기반으로 의미있는 문장을 구성하였다. 문장을 남자 화자로 하여금 읽게 하고 녹음하여 음성 DB를 구축하였다. 음성 DB에 대해 각 음소의 운율 정보를 수집하여 신경망에 맞는 목표 패턴과 훈련 패턴을 작성하였다. 이 목표 패턴은 회귀분석을 통한 추세선을 이용해 피치와 에너지에 대한 2차 다항식계수로 구성하였다. 본 논문은 목표패턴에 맞는 신경망을 학습시켜 좋은 결과를 얻었다.

• Journal of Convergence for Information Technology
• /
• v.8 no.6
• /
• pp.375-380
• /
• 2018

This paper proposes a classification method using the Convolutional Neural Network(CNN) which can obtain the type of trucks from the input image without the feature extraction step. To automatically classify vehicle images according to the type of truck cargo box, the top view images of the vehicle are used as input image and we design the structure of the CNN suitable for the input images. Learning images and correct output results is generated and the weights of neural network are obtained through the learning process. The actual image is input to the CNN and the output of the CNN is calculated. The classification performance is evaluated through comparison CNN output with actual vehicle types. Experimental results show that vehicle images could be classified with more than 90 percent accuracy according to the type of cargo box and this method can be used for pre-classification for inspecting loading defect.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.1
• /
• pp.37-43
• /
• 2020

Recently, many researches have been actively to diagnose symptoms of heart disease using ECG signal, which is an electrical signal measuring heart status. In particular, the electrocardiogram signal can be used to monitor and diagnose arrhythmias that indicates an abnormal heart status. In this paper, we proposed 1-D convolutional neural network for arrhythmias classification systems. The proposed model consists of deep 11 layers which can learn to extract features and classify 5 types of arrhythmias. The simulation results over MIT-BIH arrhythmia database show that the learned neural network has more than 99% classification accuracy. It is analyzed that the more the number of convolutional kernels the network has, the more detailed characteristics of ECG signal resulted in better performance. Moreover, we implemented a practical application based on the proposed one to classify arrythmias in real-time.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.3
• /
• pp.312-318
• /
• 2022

This paper proposes a model which classifies the type of guns and information about sound source location using deep neural network. The proposed classification model is composed of convolutional neural networks (CNN) and long short-term memory (LSTM). For training and test the model, we use the Gunshot Audio Forensic Dataset generated by the project supported by the National Institute of Justice (NIJ). The acoustic signals are transformed to Mel-Spectrogram and they are provided as learning and test data for the proposed model. The model is compared with the control model consisting of convolutional neural networks only. The proposed model shows high accuracy more than 90 %.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.15 no.6
• /
• pp.1081-1088
• /
• 2020

In this paper, the method of detecting sea fog through CCTV image is proposed based on convolutional neural networks. The study data randomly extracted 1,0004 images, sea-fog and not sea-fog, from a total of 11 ports or beaches (Busan Port, Busan New Port, Pyeongtaek Port, Incheon Port, Gunsan Port, Daesan Port, Mokpo Port, Yeosu Gwangyang Port, Ulsan Port, Pohang Port, and Haeundae Beach) based on 1km of visibility. 80% of the total 1,0004 datasets were extracted and used for learning the convolutional neural network model. The model has 16 convolutional layers and 3 fully connected layers, and a convolutional neural network that performs Softmax classification in the last fully connected layer is used. Model accuracy evaluation was performed using the remaining 20%, and the accuracy evaluation result showed a classification accuracy of about 96%.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.17 no.5
• /
• pp.55-64
• /
• 2018

In this paper, we propose fully convolutional neural networks based automatic detection of a pothole that directly causes driver's safety accidents and the vehicle damage. First, the training DB is collected through the camera installed in the vehicle while driving on the road, and the model is trained in the form of a semantic segmentation using the fully convolutional neural networks. In order to generate robust performance in a dark environment, we augmented the training DB according to brightness, and finally generated a total of 30,000 training images. In addition, a total of 450 evaluation DB was created to verify the performance of the proposed automatic pothole detection, and a total of four experts evaluated each image. As a result, the proposed pothole detection showed robust performance for missing.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.07a
• /
• pp.661-662
• /
• 2021

본 논문에서는 인공신경망 기반의 슈퍼 해상도(Super-resolution, SR) 기법을 이용하여 저해상도(Low-resolution, LR) 헤어 시뮬레이션을 고해상도(High-resolution, HR)로 노이즈 없이 표현할 수 있는 기법을 제안한다. LR과 HR 머리카락 간의 쌍은 헤어 시뮬레이션을 통해 얻을 수 있으며, 이렇게 얻어진 데이터를 이용하여 HR-LR 데이터 쌍을 설정한다. 학습할 때 사용되는 데이터는 머리카락의 위치를 지오메트리 이미지로 변환하여 사용한다. 우리가 제안하는 헤어 네트워크는 LR 이미지를 HR 이미지로 업스케일링 시키는 이미지 합성기를 위해 사용된다. 테스트 결과로 얻어진 HR 이미지가 HR 머리카락으로 다시 변환되면, 하나의 매핑 함수로 표현하기 어려운 머리카락의 찰랑거리는(Elastic) 움직임을 잘 표현할 수 있다. 합성 결과에 대한 성능으로는 전통적인 물리 기반 시뮬레이션보다 빠른 성능을 보였으며, 복잡한 수치해석을 몰라도 쉽게 실행이 가능하다.

• Journal of the Korean Society of Radiology
• /
• v.14 no.3
• /
• pp.211-220
• /
• 2020

High-quality computed tomography (CT) images enable precise lesion detection and accurate diagnosis. A lot of studies have been performed to improve CT image quality while reducing radiation dose. Recently, deep learning-based techniques for improving CT image quality have been developed and show superior performance compared to conventional techniques. In this study, a super-resolution convolutional neural network (SRCNN) model was used to improve the spatial resolution of CT images, and image quality according to the hyperparameters, which determine the performance of the SRCNN model, was evaluated in order to verify the effect of hyperparameters on the SRCNN model. Profile, structural similarity (SSIM), peak signal-to-noise ratio (PSNR), and full-width at half-maximum (FWHM) were measured to evaluate the performance of the SRCNN model. The results showed that the performance of the SRCNN model was improved with an increase of the numbers of epochs and training sets, and the learning rate needed to be optimized for obtaining acceptable image quality. Therefore, the SRCNN model with optimal hyperparameters is able to improve CT image quality.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.6
• /
• pp.629-636
• /
• 2022

We present a preprocessing method for an underwater target detection model based on a convolutional neural network. The acoustic characteristics of the ship show ambiguous expression due to the strong signal power of the low frequency. To solve this problem, we combine feature preprocessing methods with various feature scaling methods and spectrogram methods. Define a simple convolutional neural network model and train it to measure preprocessing performance. Through experiment, we found that the combination of log Mel-spectrogram and standardization and robust scaling methods gave the best classification performance.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.1
• /
• pp.75-80
• /
• 2021

Positioning technology is performing important functions in augmented reality, smart factory, and autonomous driving. Among the positioning techniques, the positioning method using beacons has been considered a challenging task due to the deviation of the RSSI value. In this study, the position of a moving object is predicted by training a neural network that takes the RSSI value of the receiver as an input and the distance as the target value. To do this, the measured distance versus RSSI was collected. A neural network was introduced to create synthetic data from the collected actual data. Based on this neural network, the RSSI value versus distance was predicted. The real value of RSSI was obtained as a neural network for generating synthetic data, and based on this value, the coordinates of the object were estimated by learning a neural network that tracks the location of a terminal in a virtual environment.