• Journal of Korea Multimedia Society
• /
• v.1 no.1
• /
• pp.26-36
• /
• 1998

It is very important study field in computer vision 'How we obtain 3D information from 2D image'. For this purpose, we must know position of camera, direction of light source, and surface reflectance property before we take the image, which are intrinsic information of the object in the scene. Among them, surface reflectance property presents very important clues. Most previous researches assume that objects have only Lambertian reflectance, but many real world objects have Non-Lambertian reflectance property. In this paper the new method for analyzing the properties of surface reflectance and reconstructing the shape of object through estimation of reflectance parameters is proposed. We have interest in Non-Lambertian reflectance surface that has specular reflection and diffuse reflection which can be explained by Torrance-Sparrow model. Photometric matching method proposed in this paper is robust method because it match reference image and object image considering the neighbor brightness distribution. Also in this thesis, the neural network based shaped reconstruction method is proposed, which can be performed in the absence of reflectance information. When brightness obtained by each light is inputted, neural network is trained by surface normal and can determine the surface shape of object.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.9C
• /
• pp.564-572
• /
• 2011

Distributed video coding (DVC) system has been proposed to reduce encoder complexity by using the correlation of frames in decoders. Since the block based motion estimation operation is not performed in the encoder of DVC system, lapped transforms, in which adjacent two blocks are transformed into one block, can be efficiently used in the DVC system. In this paper, an efficient DVC system using lapped transforms is proposed. The overlapped block motion compensated interpolation is used to produce side information, and the corresponding correlation noise between original Wyner-Ziv frame and side information is modeled. Extensive computer simulations show that the proposed DVC system outperforms conventional DVC systems.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.41 no.12
• /
• pp.1876-1883
• /
• 2016

The new channel model for high data rate common data link(CDL) is proposed. The Two-ray channel, which is composed of the reflected signals on the front ground of the receiver, is considered in this paper. This channel arises due to the curvature of the earth when the altitude of the transmitter is tens of kilometers and distance between the transmitter and the receiver is hundreds of kilometers. The Two-ray channel is modeled by estimating the maximum delay profile and the power delay profile, depending on the transmitting and receiving beamforming angle and the radiation pattern of antenna. The power delay profile has a larger effect on the bit error rate(BER) over signal to noise ratio(SNR) than the maximum delay profile, because the distance range is too long in the proposed channel model.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.2C
• /
• pp.262-271
• /
• 2004

There are channel errors when video data are transmitted between encoder and decoder. These channel errors would make decoded image incorrect, so it is very important to detect and recover channel errors. This paper proposes a method of error detection and recovery by hiding specific information into video bitstream using fragile watermark and checking it later. The proposed method requires no additional bits into compressed bitstream since it embeds a user-specific data pattern in the least significant bits of LEVELs in VLC codewords. The decoder can extract the information to check whether the received bitstream has an error or not. We also propose to use this method to embed essential data such as motion vectors that can be used for error recovery. The proposed method can detect corrupted MBs that usually escape the conventional syntax-based error detection scheme. This proposed method is quite simple and of low complexity. So the method can be applied to multimedia communication system in low bitrate wireless channel.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.447-453
• /
• 2020

In this paper, we propose a system to extract effective speaker representations from a speech signal using a deep learning method. Based on the fact that speech signal contains identity unrelated information such as text content, emotion, background noise, and so on, we perform a training such that the extracted features only represent speaker-related information but do not represent speaker-unrelated information. Specifically, we propose an auto-encoder based disentanglement method that outputs both speaker-related and speaker-unrelated embeddings using effective loss functions. To further improve the reconstruction performance in the decoding process, we also introduce a discriminator popularly used in Generative Adversarial Network (GAN) structure. Since improving the decoding capability is helpful for preserving speaker information and disentanglement, it results in the improvement of speaker verification performance. Experimental results demonstrate the effectiveness of our proposed method by improving Equal Error Rate (EER) on benchmark dataset, Voxceleb1.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.46 no.4
• /
• pp.132-138
• /
• 2009

In this paper, we propose a novel approach to improve the performance of minima controlled recursive averaging (MCRA) which is based on the second-order conditional maximum a posteriori (CMAP). From an investigation of the MCRA scheme, it is discovered that the MCRA method cannot take full consideration of the inter-frame correlation of voice activity since the noise power estimate is adjusted by the speech presence probability depending on an observation of the current frame. To avoid this phenomenon, the proposed MCRA approach incorporates the second-order CMAP criterion in which the noise power estimate is obtained using the speech presence probability conditioned on both the current observation and the speech activity decisions in the previous two frames. Experimental results show that the proposed MCRA technique based on second-order conditional MAP yields better results compared to the conventional MCRA method.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.4 no.4
• /
• pp.274-280
• /
• 2009

In order to estimate the accurate location of a user, Global Positioning system (GPS) requires at least four satellites. Since a conventional despreader operate for an GPS signal of interest, we need multiple despreaders for detecting multiple GPS signals. In this paper, we introduce the extension of the recently proposed system consisting of a null despreader, a conventional despreader, multi-stage CM (constant modulus) array, for the multiple GPS signals, and present the mathematical expression of the signal-to-interference-and-noise ratio (SINR). The extended system does not require the exact information of the direction of arrival (DOA) to suppress the directional interferences. We present the computer simulation to demonstrate the interference suppression performance of the proposed system for multiple GPS signals.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.9C
• /
• pp.906-914
• /
• 2009

SC-FDMA is employed in the 3GPP-LTE standard as the uplink transmission scheme. SC-FDMA has advantages that the signal has a low PAPR property and a simple equalizer such as FD-LE can be implemented. But FD-LE has inferior performance to Hybrid-DFE composed of frequency-domain feedforward filter and time-domain feedback filter. Recently, several IB-DFE algorithms have been proposed to overcome the disadvantages of implementation and processing complexity of Hybrid-DFE and to obtain superior performance to FD-LE. In this paper, we apply several IB-DFE algorithms to 3GPP-LTE uplink system and compare their performance by calculating BER. We investigate the effects of channel estimation errors and Doppler shift on performance. Finally, by analyzing computational complexity of IB-DFEs, we present some criteria to choose appropriate algorithm and to decide the number of iterative processes.

• Journal of Convergence for Information Technology
• /
• v.10 no.8
• /
• pp.35-39
• /
• 2020

DNN error is small compared to the conventional speech recognition system, DNN is difficult to parallel training, often the amount of calculations, and requires a large amount of data obtained. In this paper, we generate a phoneme unit to estimate the GMM parameters with each phoneme model parameters from the GMM to solve the problem efficiently. And it suggests ways to improve performance through clustering for a specific vocabulary to effectively apply them. To this end, using three types of word speech database was to have a DB build vocabulary model, the noise processing to extract feature with Warner filters were used in the speech recognition experiments. Results using the proposed method showed a 97.9% recognition rate in speech recognition. In this paper, additional studies are needed to improve the problems of improved over fitting.

• Journal of Digital Convergence
• /
• v.17 no.12
• /
• pp.211-217
• /
• 2019

Location positioning systems are available in applications such as mobile, robotic tracking systems and Wireless location-based service (LBS) applications. The GPS system is the most well-known location tracking system, but it is not easy to use indoors. The method of radio frequency identification (RFID) location tracking was studied in terms of cost-effectiveness for indoor location tracking systems. Most RFID systems use active RFID tags using expendable batteries, but in this paper, an inexpensive indoor location tracking system using passive RFID tags has been developed. A pattern matching method and a system for tracing location by generating regression curves were studied to use precision tracking algorithms. The system was tested by verifying the level of error caused by noise. The three-dimensional curves are produced by the regression equation estimated the statistically meaningful coordinates by the differential equation. The proposed system could also be applied to mobile robot systems, AGVs and mobile phone LBSs.