• Title/Summary/Keyword: codebook

Search Result 346, Processing Time 0.03 seconds

A Study on Speaker Recognition Using MFCC Parameter Space (파마메터 공간을 이용한 화자인식에 관한 연구)

  • Lee Yong-woo;Lim dong-Chol;Lee Haing Sea
    • Proceedings of the Acoustical Society of Korea Conference
    • /
    • autumn
    • /
    • pp.57-60
    • /
    • 2001
  • This paper reports on speaker-Recognition of context independence-speaker recognition in the field of the speech recognition. It is important to select the parameter reflecting the characteristic of each single person because speaker-recognition is to identify who speaks in the database. We used Mel Frequency Cesptrum Coefficient and Vector Quantization to identify in this paper. Specially, it considered to find characteristic-vector of the speaker in different from known method; this paper used the characteristic-vector which is selected in MFCC Parameter Space. Also, this paper compared the recognition rate according to size of codebook from this database and the time needed for operation with the existing one. The results is more improved $3\sim4\%$ for recognition rate than established Vector Quantization Algorithm.

  • PDF

A Speech Coder using the Simplified Multi-mode Method (단순화된 다중 모드 방법을 이용한 음성 부호화기)

  • 강홍구
    • Proceedings of the Acoustical Society of Korea Conference
    • /
    • 1995.06a
    • /
    • pp.146-149
    • /
    • 1995
  • This paper proposes a SM-CELP speech coder which applies different excitation signal according to the characteristic of speech segment at bit-rate below 4 kbps. Speech signal is divided with 2 modes such as stationary voice and etc. using the parameters of average energy of the short-time speech and the residual signal after long term prediction. Structured multi-pulse method is used for the excitation of mode-A and gaussian or pulse-like codebook for mode-B. 4.8kbps DoD-CELP are used to evaluate the performance of the proposed coder. As a result, the propose method shows 1~2 dB higher segmental signal to noise ratio and better subjectional quality without increasing the computational amount.

  • PDF

On Codebook Fesign to Improve Speaker Adaptation (화자 적응 성능 향상을 위한 코드북 설계)

  • 양태영
    • Proceedings of the Acoustical Society of Korea Conference
    • /
    • 1995.06a
    • /
    • pp.228-231
    • /
    • 1995
  • 반연속 HMM 음성인식 시스템의 화자 적응 성능 향상을 위해 코드북 변환 알고리즘을 제안하였다. 기존의 화자 적응 알고리즘으로는 새로운 화자의 적응 데이터 특징의 분포와 HMM 모수의 사전밀도를 함께 고려하는 베이시안 화자적응 알고리즘이 있다. 그러나 새로운 화자의 특징분포와 코드북 사전 밀도의 차이가 큰 경우 적응 데이터와 코드북간의 잘못된 대응 관계를 얻을 수 있으며, 기준 코드북에 필요 이상으로 많은 코드워드가 존재하는 경우 적응된 코드북에도 불필요한 코드워드 들이 남아 인식 과정에 혼란을 줄 수 있다. 이 문제점을 해결하기 위하여 제안된 코드북 변환 알고리즘에서는 주파수 영역의 포만트 정보를 이용하였다. 화자 적응을 수행하기 앞서 코드북의 켑스트럼으로부터 포만트를 추출해 내고, 이들의 분포를 적응 화자의 포만트 분포와 일치되도록 변환시켜 주었다. 이 변환된 포만트들로부터 다시 켑스트럼을 구하여 변환된 코드북을 얻고 이를 화자 적응의 초기 코드북으로 사용하였다. 제안된 알고리즘을 이용하였을 경우 코드북과 적응 화자의 음성 간의 정확한 대응관계를 찾을 수 있었고, 불필요한 코드워드들이 인식 과정에서 사용되지 않도록 변환되어 인식률이 향상되는 것을 실험을 통해 확인하였다.

  • PDF

Encoding of Speech Spectral Parameters Using Adaptive Vector-Scalar Quantization Methods for Mobile Communication Systems

  • Lee, In-Sung;Kim, Jong-Hark
    • The Journal of the Acoustical Society of Korea
    • /
    • v.17 no.4E
    • /
    • pp.35-40
    • /
    • 1998
  • In this paper, an efficient quantization method of line spectrum pairs(LSP) with cascaded structure of vector quantizer and scalar quantizer is proposed. First, input LSP parameters is vector-quantized using a codebook a with a moderate number of entries. In the second stage of quantization, the components of residual vector are individually quantized by the scalar quantizer. The utilization of ordering property of LSP parameters and the inclusion of interframe prediction improve the quantizer performance and remove the stability check routine after quantization procedure. The new vector-scalar hybrid quantizer using 26 bits/frame shows a transparent quality of speech that an average spectral distortion is 1 dB and the frame proportion with above 2 dB spectral distortion is less than 2%. The performances of proposed quantization method is evaluated in the transmission errors.

  • PDF

Fast Codebook Search Method using Triangle Inequality for Vector Quantization (백터 양자화를 위한 삼각 부등식을 이용하는 빠른 코드북 탐색법)

  • 김성재;안철웅;김승호
    • Proceedings of the Korean Information Science Society Conference
    • /
    • 1998.10c
    • /
    • pp.526-528
    • /
    • 1998
  • 영상 자료는 일반적으로 많은 정보량을 가지기 때문에 저장 공간과 전송 시간의 문제 등이 발생한다. 이 문제를 해결하기 위해 영상 압축 기법이 사용되며 그 방법 중의 하나로 벡터 양자화가 있다. 벡터 양자화는 압축률은 높지만 시간이 많이 걸리는데, 전체 처리 시간 중에서도 영상의 각 블록에 해당하는 코드벡터를 찾기 위해 주어진 코드북을 탐색하는 단계에 소요되는 시간이 가장 큰 비중을 차지한다. 본 논문에서는 코드북 탐색에 소요되는 시간을 줄여 벡터 양자화를 빠르게 하기 위한 방법으로 삼각 부등식을 이용하는 빠른 코드북 탐색법을 제안한다. 제안된 방법은 삼각 부등식을 이용해 구한 하한값을 기준으로 불필요한 계산을 줄여서 탐색 속도를 증가시킨다. 제안된 방법의 평가를 의해 100장의 256$\times$256, 256 레벨 흑백 영상을 사용하였고, 기존의 전체 탐색 방법에 비해 배 이상의 속도 향상을 얻을 수 있었다.

  • PDF

Design of PC-based CR-PACS using Multiresolution Wavelet Transform (Multiresolution Wavelet Transform을 이용한 Small PACS의 설계)

  • Kim, K.M.;Yoo, S.K.;Kim, N.H.;Huh, J.M.;Kim, E.J.
    • Proceedings of the KOSOMBE Conference
    • /
    • v.1997 no.11
    • /
    • pp.184-187
    • /
    • 1997
  • Small PACS based on a personal computer or CR has initially been designed to improve the performance of cost-effective PACS implementation. In that system, Wavelet compression scheme is newly adopted to store images hierarchically to storage unit, and retrieve and display images progressively or display workstation. In this compression method, image is decomposed into subclasses of image by wavelet transform, and then the subclasses of image are vector quantized using a multiresolution codebook.

  • PDF

An Efficient Vocoder for Digital Cellular System (디지틀 셀룰라 시스템을 위한 효율적인 음성부호화 기술)

  • 강상원;이인성;한경호
    • The Journal of Korean Institute of Communications and Information Sciences
    • /
    • v.18 no.9
    • /
    • pp.1348-1357
    • /
    • 1993
  • In this paper, we present code-excited linear prediction coding using trellis coded vector excitation, termed trellis code-excited linear prediction coding (TCELP), for an efficient 8 kbps digital cellular vocoder. A training sequence-based algorithm is developed for designing an optimized codebook subject to the TCELP structure. Also, we discuss the encoding complexity of the TCELP system and trellis symbol release rules that avoid excessive encoding delay. finally, simulation results results for the TCELP coder are given at the bit rate of 8 kbps.

  • PDF

Image coding using blocked zerotree

  • Lee, Jin-Ho;Nam, In-Gil;Park, Sang-Ho
    • Journal of Korea Society of Industrial Information Systems
    • /
    • v.6 no.1
    • /
    • pp.39-47
    • /
    • 2001
  • A blocked zerotree coding algorithm for compression of subband image is proposed. Significance of blocks with respect to a certain threshold are coded with a set of transition rules for the significance of blocks. Significant blocks are quantized by vector quantization. The basic idea for this coding approach are: 1) Subband images are coded by blocks, 2) Important blocks based on the significance of blocks are coded and quantized, 3) Multiband codebook which is composed of sub-codebooks dedicated for each threshold and subband level is adapted to produce good reproduction vectors for vector quantization. The compression results are similar to Shapiro's zerotree coding even though ours are obtained without entropy coding of bit streams from blocked zerotree encoder. If an entropy coding is applied to the bitstream, PSNR will be improved.

  • PDF

A Lossless Data Hiding Scheme for VQ Indexes Based on Joint Neighboring Coding

  • Rudder, Andrew;Kieu, The Duc
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.9 no.8
    • /
    • pp.2984-3004
    • /
    • 2015
  • Designing a new reversible data hiding technique with a high embedding rate and a low compression rate for vector quantization (VQ) compressed images is encouraged. This paper proposes a novel lossless data hiding scheme for VQ-compressed images based on the joint neighboring coding technique. The proposed method uses the difference values between a current VQ index and its left and upper neighboring VQ indexes to embed n secret bits into one VQ index, where n = 1, 2, 3, or 4. The experimental results show that the proposed scheme achieves the embedding rates of 1, 2, 3, and 4 bits per index (bpi) with the corresponding average compression rates of 0.420, 0.483, 0.545, and 0.608 bit per pixel (bpp) for a 256 sized codebook. These results confirm that our scheme performs better than other selected reversible data hiding schemes.

Fast Codebook Search for Vector Quantization in Image Coding (영상 부호화를 위한 벡터 양자화기에서의 고속 탐색 기법)

  • 고종석;김재균
    • The Journal of Korean Institute of Communications and Information Sciences
    • /
    • v.13 no.4
    • /
    • pp.302-308
    • /
    • 1988
  • The paper describes a very simple algorithm for reducing the encoding complexity of vector quantization(VQ), exploiting the feature of a vector currently being encoded. A proposed VQ of 16(=4x4) vector dimension shows a slight performance degradation of about 0.1-1.9dB, however, with only 16-32 among 256 codeword searches, i.e., with just 1/16-1/8 search complexity compared to a full-search VQ. And the proposed VQ scheme is also compared to outperform tree-search VQ with regard to their SNR performance and memory requirement.

  • PDF