• The Journal of the Acoustical Society of Korea
• /
• v.20 no.4
• /
• pp.29-34
• /
• 2001

In this paper, we designed an LSF coefficient quantizer of the wideband speech codec that can produce high quality speech service. For the efficient LSF coefficient quantizer, the interframe correlation was used. Also we separately quantized the LSF coefficients with high and low interframe correlation. Predictive pyramid vector quantizer (PVQ) was used for quantizing the LSF coefficients with high interframe correlation, and PVQ was used for quantizing the LSF coefficients with low interframe correlation. Experiments show that the proposed UF quantizer can quantize LSF information in 40 bits/frame, with an average spectral distortion (SD) of 1 dB and less than 3.87% frames having SD greater than 2 dB.

• The Journal of the Acoustical Society of Korea
• /
• v.16 no.1
• /
• pp.10-15
• /
• 1997

In this paper, an efficient LSF quantizer, named improved PSVQ(IPSVQ), is proposed to apply in the 8 kbps QCELP speech coder. By using 27 bits IPSVQ instead of 40 bits DPCM quantizer per frame, we can save 13 bits/frame and allocate those bits to the codebook gain and the pitch gain parameters. Hence we improve the overall performance of the QCELP codec. The enhanced QCELP shows the performance improvement of 0.9 dB SNR and 0.4 dB SEGSNR. Informal listening tests also confirm the improvement in the speech quality.

• The Journal of the Acoustical Society of Korea
• /
• v.31 no.2
• /
• pp.93-99
• /
• 2012

For speech coding, a vocal tract is modeled using Linear Predictive Coding (LPC) coefficients. The LPC coefficients are typically transformed to Line Spectral Frequency (LSF) parameters which are advantageous for linear interpolation and quantization. If multidimensional LSF data are quantized directly using Vector-Quantization (VQ), high rate-distortion performance can be obtained by fully utilizing intra-frame correlation. In practice, since this direct VQ system cannot be used due to high computational complexity and memory requirement, Split VQ (SVQ) is used where a multidimensional vector is split into multilple sub-vectors for quantization. The LSF parameters also have high inter-frame correlation, and thus Predictive SVQ (PSVQ) is utilized. PSVQ provides better rate-distortion performance than SVQ. In this paper, to implement the optimal predictors in PSVQ for voice storage devices, we propose Multi-Frame AR-model based SVQ (MF-AR-SVQ) that considers the inter-frame correlations with multiple previous frames. Compared with conventional PSVQ, the proposed MF-AR-SVQ provides 1 bit gain in terms of spectral distortion without significant increase in complexity and memory requirement.

• Proceedings of the IEEK Conference
• /
• 1999.06a
• /
• pp.679-682
• /
• 1999

This paper proposes the LSF or LSP that is the method of using to transfer the speech parameters after processed the speech to LPC, which is digital coding transferring efficiently, for the best quality and the lowest bit rate of parameters. The new revised transform algorithm between LSF and LPC coefficients is proposed. The proposed algorithm eliminates all multiplications, computes fewer operations, and reduces memory buffer sizes.

• Proceedings of the Acoustical Society of Korea Conference
• /
• spring
• /
• pp.153-156
• /
• 2000

본 논문에서는 멀티미디어 서비스들 중에서 음성 또는 오디오 신호를 저속으로 압축할 때 사용되는 G.723.1 부호화기의 line spectral frequency(LSF) 계수 양자화 방식을 고속으로 처리하는 알고리즘을 제안하였다. 제안된 고속탐색 방법은 LSF 계수의 순서성질을 이용하여 코드북의 탐색 범위를 줄임으로써 계산량을 크게 감소시킨다. 제안된 고속탐색 방법을 predictive split VQ(PSVQ) 구조를 갖는 G.723.1 에 적용한 결과 spectral distortion(SD) 성능 감쇄 및 추가적인 메모리 증가 없이 최적 코드벡터를 찾기 위한 코드북 탐색 과정에서 코드북의 평균 탐색 범위가 $20.1\%$ 감소했으며, 이는 additions, subtractions, multiplies 및 comparisons 수가 각각 $19.1\%$, $20.1\%$, $19.4\%$ 및 $12.2\% 감소하는 결과를 얻었다.

• The Journal of the Acoustical Society of Korea
• /
• v.23 no.4
• /
• pp.333-339
• /
• 2004

This Paper proposes the vector quantizer-pyramid vector quantizer(VQ-PVQ) structure. Also both predictive structure and safety-net concept are combined into the VQ-PVQ to quantize the IPC parameter of wideband speech codec. The Performance is compared to the LPC vector quantizer used in the AMR-WB(ITU-T G.722.2). demonstrating reduction in both spectral distortion and encoding memory.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.3C
• /
• pp.234-240
• /
• 2007

In this paper, low complexity block constrained trellis coded quantization (BC-TCQ) structures are introduced, and a predictive BC TCQ encoding method is developed for quantization of line spectrum frequencies (LSF) parameters for narrowband speech coding applications. Trellis-coded quantization(TCQ) is a form of VQ that builds the VQ codebook from interleaved constituent scalar quantization codebooks. The performance is compared to the other VQ, demonstrating reduction in spectral distortion and significant reduction in encoding complexity. The predictive BC-TCQ is about 0.47107 dB superior to the IS-641 split-VQ, 26bits/frame, in spectral distortion sense. The BC-TCQ is 64.54%, 76.93%, 2.35% of the IS-641 split-VQ, respectively, in the complexity of the additions, multiplies, comparisons.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.3-6
• /
• 2000

CELP 계열 보코더중 인터넷 폰 및 화상회의를 목적으로 개발된 G.723.1 보코더에서는 LPC를 LSP로 변환하기 위해 LPC 계수를 이용한 다항식을 구성한 다음 근을 검색하는 방법을 사용하고 있다. 근 검색시에는 256/pi의 범위 안에서 동일한 간격을 갖는 코사인 테이블을 구성하여 순차적으로 검색하게 된다. LSF의 근들은 포만트가 존재하는 대역에서 근들이 나타나게 되므로 유성음의 경우 저주파수 대역에서 무성음의 경우 고주파수 대역에서 많이 분포하게 된다. 하지만 G.723.1에서 사용하는 코사인 테이블은 음성신호의 특성을 고려하지 않고 균등한 간격을 갖는 값들을 사용함으로 음질을 저해할 수 있는 요소를 갖고 있다. 따라서 본 논문에서는 음성의 특성을 고려한 코사인 테이블을 재구성함으로써 음질을 향상시킬 수 있었으며 주관적 음질평가인 MOS 시험결과 평균 1.8 정도의 음질향상을 가져올 수 있었다

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.26 no.1
• /
• pp.33-40
• /
• 2014

Reliability analysis of jacket type offshore wind turbine (OWT) support structure under extreme ocean environmental loads was performed. Limit state function (LSF) of OWF support structure is defined by using structural dynamic response at mud-line. Then, the dynamic response is expressed as the static response multiplied by dynamic response factor (DRF). Probabilistic distribution of DRF is found from response time history under design significant wave load. Band limited beta distribution is used for internal friction angle of ground soil. Wind load is obtained in the form of thrust force from commercial code called GH_Bladed and then, applied to tower hub as random load. In a numerical example, the response surface method (RSM) is used to express LSF of jacket type support structure for 5MW OWF. Reliability index is found using first order reliability method (FORM).

• Journal of Korea Water Resources Association
• /
• v.48 no.6
• /
• pp.491-500
• /
• 2015

In order to improve the runoff estimation accuracy of the Natural Resources Conservation Service (NRCS) curve number (CN) model, this study incorporated rainfall and maximum potential retention as contributors for initial abstraction. The modification was proposed based on 658 rank-order data of rainfall and subsequent runoff from 15 watersheds. The NRCS-CN model (M1), one of its inspired modified model (M2), and the proposed model (M3) were analyzed employing different CN approaches. Using tabulated, calculated and least squares fitted CNs ($CN_T$, $CN_C$, $CN_{LSF}$, respectively), the models' performances were evaluated based on Root Mean Square Error (RMSE), Nash-Sutcliffe Efficiency (NSE), and Percent Bias (PBIAS). Applications of model M1, M2, and M3, respectively exhibited watershed cumulative mean [RMSE (23.60, 18.12, 16.04), NSE (0.54, 0.73, 0.79), and PBIAS (36.54, 20.25, 12.00)]. Similarly, using CNC (for M1 and M2 model) and $CN_{LSF}$ (for M3 model), the performance of three models respectively were assessed based on watershed cumulative mean [RMSE (17.17, 15.88, 13.82), NSE (0.76, 0.80, 0.85), and PBIAS (3.06, 4.47, 0.11)]. The proposed model (M3) that linked all of the NRCS-CN variants showed more statistically significant agreement between the observed and estimated data.