The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Exibições de texto completo
91
Formulamos o projeto de um quantizador ótimo como um problema de otimização que encontra os índices de quantização que minimizam o erro de quantização. Como solução do problema de otimização, é proposta uma abordagem baseada em programação dinâmica, chamada de quantização DP. Observa-se que os sinais quantizados nem sempre contêm todos os tipos de valores de sinal que podem ser representados com determinada profundidade de bits. Esta propriedade é chamada de dispersão de amplitude. Como a quantização é a discretização da amplitude do valor do sinal, a dispersão da amplitude está intimamente relacionada ao projeto do quantizador. Valores de sinal com frequência zero não afetam o erro de quantização, portanto, existe o potencial de reduzir a complexidade do quantizador ideal ao não calcular valores de sinal com frequência zero. No entanto, os métodos convencionais para quantização de DP não foram projetados para considerar a dispersão de amplitude e, portanto, não conseguem reduzir a complexidade. O algoritmo proposto oferece um quantizador ideal de complexidade reduzida que minimiza o erro de quantização enquanto aborda a dispersão de amplitude. Os resultados experimentais mostram que o algoritmo proposto pode alcançar uma redução de complexidade em relação à quantização DP convencional em 82.9 a 84.2%, em média.
Yukihiro BANDOH
NTT Corporation
Seishi TAKAMURA
NTT Corporation
Atsushi SHIMIZU
NTT Corporation
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copiar
Yukihiro BANDOH, Seishi TAKAMURA, Atsushi SHIMIZU, "Sparse DP Quantization Algorithm" in IEICE TRANSACTIONS on Fundamentals,
vol. E102-A, no. 3, pp. 553-565, March 2019, doi: 10.1587/transfun.E102.A.553.
Abstract: We formulate the design of an optimal quantizer as an optimization problem that finds the quantization indices that minimize quantization error. As a solution of the optimization problem, an approach based on dynamic programming, which is called DP quantization, is proposed. It is observed that quantized signals do not always contain all kinds of signal values which can be represented with given bit-depth. This property is called amplitude sparseness. Because quantization is the amplitude discretization of signal value, amplitude sparseness is closely related to quantizer design. Signal values with zero frequency do not impact quantization error, so there is the potential to reduce the complexity of the optimal quantizer by not computing signal values that have zero frequency. However, conventional methods for DP quantization were not designed to consider amplitude sparseness, and so fail to reduce complexity. The proposed algorithm offers a reduced complexity optimal quantizer that minimizes quantization error while addressing amplitude sparseness. Experimental results show that the proposed algorithm can achieve complexity reduction over conventional DP quantization by 82.9 to 84.2% on average.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/transfun.E102.A.553/_p
Copiar
@ARTICLE{e102-a_3_553,
author={Yukihiro BANDOH, Seishi TAKAMURA, Atsushi SHIMIZU, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Sparse DP Quantization Algorithm},
year={2019},
volume={E102-A},
number={3},
pages={553-565},
abstract={We formulate the design of an optimal quantizer as an optimization problem that finds the quantization indices that minimize quantization error. As a solution of the optimization problem, an approach based on dynamic programming, which is called DP quantization, is proposed. It is observed that quantized signals do not always contain all kinds of signal values which can be represented with given bit-depth. This property is called amplitude sparseness. Because quantization is the amplitude discretization of signal value, amplitude sparseness is closely related to quantizer design. Signal values with zero frequency do not impact quantization error, so there is the potential to reduce the complexity of the optimal quantizer by not computing signal values that have zero frequency. However, conventional methods for DP quantization were not designed to consider amplitude sparseness, and so fail to reduce complexity. The proposed algorithm offers a reduced complexity optimal quantizer that minimizes quantization error while addressing amplitude sparseness. Experimental results show that the proposed algorithm can achieve complexity reduction over conventional DP quantization by 82.9 to 84.2% on average.},
keywords={},
doi={10.1587/transfun.E102.A.553},
ISSN={1745-1337},
month={March},}
Copiar
TY - JOUR
TI - Sparse DP Quantization Algorithm
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 553
EP - 565
AU - Yukihiro BANDOH
AU - Seishi TAKAMURA
AU - Atsushi SHIMIZU
PY - 2019
DO - 10.1587/transfun.E102.A.553
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E102-A
IS - 3
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - March 2019
AB - We formulate the design of an optimal quantizer as an optimization problem that finds the quantization indices that minimize quantization error. As a solution of the optimization problem, an approach based on dynamic programming, which is called DP quantization, is proposed. It is observed that quantized signals do not always contain all kinds of signal values which can be represented with given bit-depth. This property is called amplitude sparseness. Because quantization is the amplitude discretization of signal value, amplitude sparseness is closely related to quantizer design. Signal values with zero frequency do not impact quantization error, so there is the potential to reduce the complexity of the optimal quantizer by not computing signal values that have zero frequency. However, conventional methods for DP quantization were not designed to consider amplitude sparseness, and so fail to reduce complexity. The proposed algorithm offers a reduced complexity optimal quantizer that minimizes quantization error while addressing amplitude sparseness. Experimental results show that the proposed algorithm can achieve complexity reduction over conventional DP quantization by 82.9 to 84.2% on average.
ER -