• Journal of the Korean Institute of Telematics and Electronics
• /
• v.22 no.6
• /
• pp.41-45
• /
• 1985

A suboptimum block quantization method is investigated for efficent transform coding. In our study the following method has appeared as suboptimum. At first, optimum bit allocation is done assuming the varances of the transform coefficients are known. Secndly, a varance estimation algorithm which results from the approximate equations governing the optimum vit alllocation is applied. The better performance of the proposed block quantization method has been confirmed by simulations based on varous pdf assumptions. the results indicate that the proposed method yields overall improvements of about 25% in NMSE for both the symmetric nonuniform and uniform quantizer at the coding rate of 1 bit/pel.

• The Journal of the Acoustical Society of Korea
• /
• v.23 no.6
• /
• pp.481-487
• /
• 2004

In this paper. the structure of bandwidth scalable wideband speech codec and its high-band codec are proposed. In the high-band codec. the signal is divided into frequency bands. and each band is quantized in DCT domain. The DCT coefficients are splitted into magnitude and sign, and each is quantized independently by a specialized method based on its characteristics. In addition. the quantized gain parameter in the low-band codec is utilized in the high-band codec for an enhanced performance. The bandwidth scalable wideband speech codec using G.729E for low-band and the proposed codec for high-band is developed, and it is confirmed that the proposed codec has better subjective performance than 24kbps G.722.1.

• The Journal of the Acoustical Society of Korea
• /
• v.24 no.7
• /
• pp.402-409
• /
• 2005

In this paper. we propose a coding method using a matching pursuit algorithm in a strongly periodic highband signal. Also. we propose an efficient quantizer for the estimated parameters : spectral magnitude and phase. Based on the error concealment principle and sinusoidal model. the MP algorithm requires the high-precision pitch period estimation. To estimate more accurate pitch period. the refined pitch obtained from lowband speech is used. which increases the efficiency of bit allocation. The spectral magnitude parameters are quantized by the method which is combined with MDCT (Modified Discrete Cosine Transform) and multi-stage structure. The spectral phase quantizer uses the $2{\pi}$ modular characteristic of phases and the weighted function by spectral magnitudes. To evaluate the efficiency of the proposed method. we applied it to analysis-by-synthesis system. Furthermore we suggest the possibillity of scalable wideband speech codecs based on band-split structure.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.19 no.1
• /
• pp.57-62
• /
• 2019

This paper evaluates the QE-MMA (Quantized Error-MMA) adaptive equalization algorithm by the number of quantizer in order to compensates the intersymbol interference due to channel in the transmission of high spectral efficient nonconstant modulus signal. In the adaptive equalizer, the error signal is needed for the updating the tap coefficient, the QE-MMA uses the polarity of error signal and correlation multiplier that condered nonlinear finite bit power-of-two quantizing component in order to convinience of H/W implementation. The different adaptive equalization performance were obtained by the number of quantizer, these performance were evaluated by the computer simulation. For this, the equalizer output signal constellation, residual isi, maximum distortion, MSE, SER were applied as a performance index. As a result of computer simulation, it improved equalization performance and reduced equalization noise were obtained in the steady state by using large quantizer bit numbers, but gives slow in convergence speed for reaching steady state.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.81-84
• /
• 2004

패킷망을 통한 음성 통신이 발달됨에 따라 패킷 스위칭 채널 환경에서 계층적 구조를 가지는 광대역 음성 부호화기의 개발에 대한 요구가 늘어나고 있다. 본 논문에서는 이러한 차세대 패킷 전화용 광대역 음성 부호화기의 상위 대역에 대해서 효율적인 양자화 방법을 제안한다. 먼저 전체 프레임을 다수의 짧은 부프레임으로 구분하고, 각각의 부프레임에 MLT(Modulated Lapped Transform)변환을 적용하여 주파수 영역으로 변환하여 2차원 구조의 데이터 행렬을 생성한다. 이러한 2차원 구조의 데이터를 크기와 부호로 분리하고, 크기는 2차원 DCT를 사용하여 시간과 주파수 영역에서의 신호 압축을 동시에 얻을 수 있게 하였다. 이와 같은 새로운 구조를 활용하여 기존의 방법보다 Energy Compaction 효과를 높이고 양자화 성능을 향상시킬 수 있었다. 또한 Core Layer의 부호화된 파라미터를 상위 대역의 양자화에 이용함으로써 그 성능을 향상시킬 수 있는 방법을 제안한다.

• The KIPS Transactions:PartB
• /
• v.12B no.4 s.100
• /
• pp.437-442
• /
• 2005

For the purpose of speaker normalization in speaker independent speech recognition systems, experiments are conducted on a method based on Gaussian mixture model(GMM). The method, which is an improvement of the previous study based on vector quantizer, consists of modeling the probability distribution of canonical feature vectors by a GMM with an appropriate number of clusters, and of estimating the warp factor of a test speaker by making use of the obtained probabilistic model. The purpose of this study is twofold: improving the existing ML based methods, and comparing the performance of what is called 'soft decision' method with that of the previous study based on vector quantizer. The effectiveness of the proposed method is investigated by recognition experiments on the TIMIT corpus. The experimental results showed that a little improvement could be obtained tv adjusting the number of clusters in GMM appropriately.

• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.469-472
• /
• 2001

MPEC Audio Layer-III(MP3)알고리듬은 복호화기에 비해 부호화기가 월등히 많은 연산량을 가지고 있는 비대칭 구조를 가지고 있다. MP3 부호화기의 대부분의 연산량은 복잡한 초월함수 연산이 포함되는 심리음향모델과 반복 루프 과정을 수행하는 비선형 양자화와 비트 할당과정 이 차지한다. 본 논문에서는 MP3 부호화기의 실시간 구현을 위한 알고리듬 레벨의 최적화를 수행하였다. MP3 부호화기의 연산량을 줄이기 위해 심리음향모델을 간략화하고 반복 루프의 회수를 최소화할 수 있는 방법을 제안하였다. 프레임당 한 그래뉼의 심리음향모델 정보를 계산하여 한 프레임 내에서의 심리음향모델 정보를 추정함으로써 연산량을 45% 이상 감소시켰다. 또한 외부 반복 루프의 반복 회수를 줄이기 위하여 외부 반복 루프의 반복에 따른 스케일 팩터(Scale Factor) 및 양자화 스탭의 증가 패턴을 관찰하고 최적화된 스캐일 팩터 증가 방법을 제안하였다. 제안된 고속화 방법은 주관적 음질 평가를 통해 성능을 검증하였다.

• The Transactions of the Korea Information Processing Society
• /
• v.3 no.4
• /
• pp.938-946
• /
• 1996

Hierarchical subband codec has been widely promoted in the field of data compression/decompression because of their simplicity and modular nature. Over the past years, the study has received great attention to the perfect reconstruction (PR)system which perfectly recovers the original input signal at the reconstructed output. However, in the actual subband codec system, the signals that passed through the analysis filter bank are quantized before transmission to the receiver side and reconstructed by the synthesis filter bank. Thus the PR system is impossible and the quantization effects must be carefully considered in the system design such that the system recovers the reconstructed output as possible to the the original input signal with minimum quantization error.In this paper, we propose an optimum hierarchical subband codec structure in the presence of quantizer. The optimality criteria of the code is given to the deign of the hierarchical analysis/synthesis subband filter bank and the quantizer that minimize then output mean square error due to the quantizer in the codec. Specific opti-mum design esamples are shown with level-1, level-2 hierarchical structure. The optimal designs are verified by computer simulation.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.7
• /
• pp.33-38
• /
• 1999

In this paper, we combine a generalized analysis-by-synthesis (AbS) structure and an algebraic excitation scheme to propose a new 4kb/s speech codec. This codec partly uses the structure of G.729. We design a line spectrum pair (LSP) quantizer, an adaptive codebook, and an excitation codebook to fit the 4 kb/s bit rate. The codec has a 25㎳ algorithmic delay, which corresponds to a 20㎳ frame size and a 5㎳ lookahead. At the bit rates below 4kb/s, most CELP speech codecs using the AbS principle have a drawback that results a rapid degradation of speech quality. To overcome this drawback we use the generalized AbS structure which is efficient for the low bit rate speech codec. LP coefficients are converted to LSP and quantized using a predictive 2-stage VQ. A low complexity algebraic codebook which uses shifting method is used for the fixed codebook excitation, and gains of the adaptive codebook and the fixed codebook are quantized using the VQ. To evaluate the performance of the proposed codec A-B preference tests are done with the fixed rate 8kb/s QCELP. As the result of the test, the performance of the codec is similar to that of the fixed rate 8kb/s QCELP.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.42 no.2 s.302
• /
• pp.49-58
• /
• 2005

In this paper, we proposed a blind watermarking algerian to use characteristics of a scalar quantizer which is the recommended in the JPEG2000 and JPEG. The proposed algorithm shifts a quantization index according to the value of each watermark bit to prevent losing the watermark information during the compression by quantization. Therefore, the watermark is embedded during the process of quantization, not an additional process for watermarking, and is adaptively applied as a assigned quantizer according application areas. Before embedding process, a LFSR(Linear feedback shift register) rearranged the watermark for the security of the watermark itself and in the embedding process, a LFSR is used to hide the watermarking positions. Therefore the embedded watermark can he extracted by only the owner who knows the initial value of LFSR without the original image. The visual recognizable pattern such as a binary image was used as the watermark. The experimental results showed that the proposed algerian satisfies the robustness and imperceptibility corresponding to the major requirement of watermarking. The results showed the largest error rate to be $5.7\%$ for attack. The experimental result which compares the proposed algorithm with the Mohamed algorithm showed that the proposed algorithm was better than it, exactly $4\~5$ times for the attacks of JPEG and JPEG2000.