• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.8 no.2
• /
• pp.352-358
• /
• 2004

This paper is studied the improvement of performance for moving picture encoder using H.263. This is used the new motion vector search algorithm using a relation with neighborhood search point and is applied the integer DCT for the encoder. The integer DCT behaves DCT by the addition operation of the integer using WHT and a integer lifting than conventional DCT that needs the multiplication operation of a floating point number. Therefore, the integer Dn can reduce the operation amount than basis DCT with having an equal PSNR. The new motion vector search algorithm is showed almost similar PSNR as reducing the operation amount than the conventional motion vector search algorithm. To experiment a compatibility of the integer DCT and the conventional DCT, according to result compare case that uses a method only and case that uses the alternate two methods of the integer DCT or the conventional DCT to H.263 encoder and decoder, case that uses the alternate two methods is showed doing not deteriorate PSNR-and being each other compatible visually than case that uses an equal method only.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.2
• /
• pp.279-285
• /
• 2011

The next generation wireless communication systems are required high BER performance better than present performance. Double binary Turbo code have error floor at high SNR, so it cannot be used in next generation wireless communication system. Therefore, many methods are proposed for overcome error floor at DVB-RCS NG(next generation). In this paper, we analysis structure of third-dimension Turbo code(3D-turbo code). 3D-Turbo code overcomes error flow by additive post-encoder in conventional DVB-RCS Turbo code. Performance of 3D-Turbo code is changed by post-encoder form, interleaving method, value of ${\lambda}$. So we are simulated by those parameter and proposed optimal form. By a result, performance of 3D-Turbo is better than conventional DVB-RCS Turbo code and it overcome error floor of conventional DVB-RCS Turbo code.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.11
• /
• pp.2015-2020
• /
• 2007

This paper presents the SVC(Scalable Video Coding) scheme which enables the AT-DMB(Advanced Terrestrial-DMB) video service in enhancement layer, while keeping the current T-DMB video service in base layer. But, it is very complicate to implement the SVC encoder and so it is necessary to analyze the complexity and performance for SVC encoder#s structures and coding parameters. In this paper, through computer simulations, SVC coding parameters are tested and then, based on these results, three types of SVC encoder models are compared from the viewpoint of the complexity and performance.

• Journal of the Korea Society of Computer and Information
• /
• v.18 no.8
• /
• pp.53-64
• /
• 2013

The existing research on the QoS of the Streaming video system(SVS) adjust the video stream being sent from the network perspective, the focus is to prevent data loss and delay. The other hand, changing the quality of the video stream will not receive the guaranteed QoS from the perspective of the users who want uniform quality. In order to solve these problems, this paper applied to ensure that the benefits in the amount of bits per frame occurs Encoder CBR streaming video is kept constant and uniform picture quality advantages of VBR streaming video VBR to CBR Encoder and CBR to VBR Decoder, Video was designed to control the playback stream for And to ensure a uniform quality of the user based on the design and implementation of uniform quality guaranteed SVS. PSNR evaluated for several characteristics of the sample video to demonstrate the superiority of the SVS ensure uniform quality, the proposed Performance evaluation of the SVS ensure uniform quality CBW the ABR from 100% to ensure uniform image quality from 85% to environmental excellence is proved.

• Journal of Internet Computing and Services
• /
• v.22 no.1
• /
• pp.13-22
• /
• 2021

Recently network based attack technologies are rapidly advanced and intelligent, the limitations of existing signature-based intrusion detection systems are becoming clear. The reason is that signature-based detection methods lack generalization capabilities for new attacks such as APT attacks. To solve these problems, research on machine learning-based intrusion detection systems is being actively conducted. However, in the actual network environment, attack samples are collected very little compared to normal samples, resulting in class imbalance problems. When a supervised learning-based anomaly detection model is trained with such data, the result is biased to the normal sample. In this paper, we propose to overcome this imbalance problem through One-Class Anomaly Detection using an auto encoder. The experiment was conducted through the NSL-KDD data set and compares the performance with the supervised learning models for the performance evaluation of the proposed method.

• Journal of Software Assessment and Valuation
• /
• v.17 no.2
• /
• pp.47-57
• /
• 2021

Neuromorphic technology is proposed to complement the shortcomings of existing artificial intelligence technology by mimicking the human brain structure and computational process with hardware. NA-IDE has also been proposed for developing neuromorphic hardware-based IoT applications. To implement an SNN model in NA-IDE, commonly used input data must be transformed for use in the SNN model. In this paper, we implemented a neural coding method encoder component that converts image data into a spike train signal and uses it as an SNN input. The decoder component is implemented to convert the output back to image data when the SNN model generates a spike train signal. If the decoder component uses the same parameters as the encoding process, it can generate static data similar to the original data. It can be used in fields such as image-to-image and speech-to-speech to transform and regenerate input data using the proposed encoder and decoder.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.3
• /
• pp.351-358
• /
• 2022

Recently, Generative Adversarial Networks (GAN) and Variational AutoEncoders (VAE) have been applied to voice conversion that can make use of non-parallel training data. Especially, Conditional Cycle-Consistent Generative Adversarial Networks (CC-GAN) and Cycle-Consistent Variational AutoEncoders (CycleVAE) show promising results in many-to-many voice conversion among multiple speakers. However, the number of speakers has been relatively small in the conventional voice conversion studies using the CC-GANs and the CycleVAEs. In this paper, we extend the number of speakers to 100, and analyze the performances of the many-to-many voice conversion methods experimentally. It has been found through the experiments that the CC-GAN shows 4.5 % less Mel-Cepstral Distortion (MCD) for a small number of speakers, whereas the CycleVAE shows 12.7 % less MCD in a limited training time for a large number of speakers.

• The Journal of the Convergence on Culture Technology
• /
• v.8 no.4
• /
• pp.91-98
• /
• 2022

Lifelog data continuously collected through a wearable device may contain many outliers, so in order to improve data quality, it is necessary to find and remove outliers. In general, since the number of outliers is less than the number of normal data, a class imbalance problem occurs. To solve this imbalance problem, we propose a method that applies Variational AutoEncoder to outliers. After preprocessing the outlier data with proposed method, it is verified through a number of machine learning models(classification). As a result of verification using body weight data, it was confirmed that the performance was improved in all classification models. Based on the experimental results, when analyzing lifelog body weight data, we propose to apply the LightGBM model with the best performance after preprocessing the data using the outlier processing method proposed in this study.

• The Journal of the Acoustical Society of Korea
• /
• v.42 no.2
• /
• pp.94-101
• /
• 2023

A deep neural network composed of encoders and decoders, such as U-Net, used for speech enhancement, concatenates the encoder to the decoder through skip-connection. Skip-connection helps reconstruct the enhanced spectrum and complement the lost information. The features of the encoder and the decoder connected by the skip-connection are incompatible with each other. In this paper, for complex-valued spectrum based speech enhancement, Self-Attention (SA) method is applied to skip-connection to transform the feature of encoder to be compatible with the features of decoder. SA is a technique in which when generating an output sequence in a sequence-to-sequence tasks the weighted average of input is used to put attention on subsets of input, showing that noise can be effectively eliminated by being applied in speech enhancement. The three models using encoder and decoder features to apply SA to skip-connection are studied. As experimental results using TIMIT database, the proposed methods show improvements in all evaluation metrics compared to the Deep Complex U-Net (DCUNET) with skip-connection only.

• Smart Media Journal
• /
• v.2 no.2
• /
• pp.39-43
• /
• 2013

In this paper, we suggest a new technique for WPC parity-check matrix (H-matrix) generation and a corresponding decoding process. The key idea is to construct WPC H-matrix by using a convolutional encoder. It is easy to have many different coderates from a mother code with convolutional codes. However, it is difficult to have many different coderates with LDPC codes. Constructing LDPC Hmatrix based on a convolutional code can easily bring the advantage of convolutional codes to have different coderates. Moreover, both LDPC and convolutional decoding algorithms can be applied altogether in the decoding part. This process prevents the performance degradation of short-length WPC code.