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
A análise de Predição Linear (LP) é comumente usada no processamento de fala. LP é baseado no modelo Auto-Regressivo (AR) e estima o parâmetro do modelo AR a partir de sinais com l2-otimização de norma. Recentemente, a estimativa esparsa recebeu atenção, pois pode extrair recursos significativos de big data. A estimativa esparsa é realizada por l1 or l0-otimização ou regularização de normas. Métodos de análise de LP esparsos baseados em l1otimização de norma foi proposta. Como a excitação da fala não é gaussiana branca, uma estimativa esparsa de LP pode estimar parâmetros mais precisos do que a estimativa convencional l2LP baseado em normas. Estas são análises invariantes no tempo e com valor real. Estudamos a análise de AR Complexo Variável no Tempo (TV-CAR) para um sinal analítico e avaliamos o desempenho no processamento de fala. Os métodos TV-CAR são l2-métodos normativos. Neste artigo, propomos a análise TV-CAR esparsa baseada no LASSO adaptativo (operador de menor encolhimento e seleção absoluta) que é l1-normizar a regularização e avaliar o desempenho em F0 estimativa da fala usando IRAPT (Instantaneous RAPT). Os resultados experimentais mostram que os métodos esparsos de TV-CAR apresentam melhor desempenho para um alto nível de ruído rosa aditivo.
Keiichi FUNAKI
University of the Ryukyus
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
Keiichi FUNAKI, "Sparse Time-Varying Complex AR (TV-CAR) Speech Analysis Based on Adaptive LASSO" in IEICE TRANSACTIONS on Fundamentals,
vol. E102-A, no. 12, pp. 1910-1914, December 2019, doi: 10.1587/transfun.E102.A.1910.
Abstract: Linear Prediction (LP) analysis is commonly used in speech processing. LP is based on Auto-Regressive (AR) model and it estimates the AR model parameter from signals with l2-norm optimization. Recently, sparse estimation is paid attention since it can extract significant features from big data. The sparse estimation is realized by l1 or l0-norm optimization or regularization. Sparse LP analysis methods based on l1-norm optimization have been proposed. Since excitation of speech is not white Gaussian, a sparse LP estimation can estimate more accurate parameter than the conventional l2-norm based LP. These are time-invariant and real-valued analysis. We have been studied Time-Varying Complex AR (TV-CAR) analysis for an analytic signal and have evaluated the performance on speech processing. The TV-CAR methods are l2-norm methods. In this paper, we propose the sparse TV-CAR analysis based on adaptive LASSO (Least absolute shrinkage and selection operator) that is l1-norm regularization and evaluate the performance on F0 estimation of speech using IRAPT (Instantaneous RAPT). The experimental results show that the sparse TV-CAR methods perform better for a high level of additive Pink noise.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/transfun.E102.A.1910/_p
Copiar
@ARTICLE{e102-a_12_1910,
author={Keiichi FUNAKI, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Sparse Time-Varying Complex AR (TV-CAR) Speech Analysis Based on Adaptive LASSO},
year={2019},
volume={E102-A},
number={12},
pages={1910-1914},
abstract={Linear Prediction (LP) analysis is commonly used in speech processing. LP is based on Auto-Regressive (AR) model and it estimates the AR model parameter from signals with l2-norm optimization. Recently, sparse estimation is paid attention since it can extract significant features from big data. The sparse estimation is realized by l1 or l0-norm optimization or regularization. Sparse LP analysis methods based on l1-norm optimization have been proposed. Since excitation of speech is not white Gaussian, a sparse LP estimation can estimate more accurate parameter than the conventional l2-norm based LP. These are time-invariant and real-valued analysis. We have been studied Time-Varying Complex AR (TV-CAR) analysis for an analytic signal and have evaluated the performance on speech processing. The TV-CAR methods are l2-norm methods. In this paper, we propose the sparse TV-CAR analysis based on adaptive LASSO (Least absolute shrinkage and selection operator) that is l1-norm regularization and evaluate the performance on F0 estimation of speech using IRAPT (Instantaneous RAPT). The experimental results show that the sparse TV-CAR methods perform better for a high level of additive Pink noise.},
keywords={},
doi={10.1587/transfun.E102.A.1910},
ISSN={1745-1337},
month={December},}
Copiar
TY - JOUR
TI - Sparse Time-Varying Complex AR (TV-CAR) Speech Analysis Based on Adaptive LASSO
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 1910
EP - 1914
AU - Keiichi FUNAKI
PY - 2019
DO - 10.1587/transfun.E102.A.1910
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E102-A
IS - 12
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - December 2019
AB - Linear Prediction (LP) analysis is commonly used in speech processing. LP is based on Auto-Regressive (AR) model and it estimates the AR model parameter from signals with l2-norm optimization. Recently, sparse estimation is paid attention since it can extract significant features from big data. The sparse estimation is realized by l1 or l0-norm optimization or regularization. Sparse LP analysis methods based on l1-norm optimization have been proposed. Since excitation of speech is not white Gaussian, a sparse LP estimation can estimate more accurate parameter than the conventional l2-norm based LP. These are time-invariant and real-valued analysis. We have been studied Time-Varying Complex AR (TV-CAR) analysis for an analytic signal and have evaluated the performance on speech processing. The TV-CAR methods are l2-norm methods. In this paper, we propose the sparse TV-CAR analysis based on adaptive LASSO (Least absolute shrinkage and selection operator) that is l1-norm regularization and evaluate the performance on F0 estimation of speech using IRAPT (Instantaneous RAPT). The experimental results show that the sparse TV-CAR methods perform better for a high level of additive Pink noise.
ER -