Maximum Phase Modeling for Sparse Linear Prediction of Speech
Linear prediction (LP) is an ubiquitous analysis method in speech processing. Various studies have focused on sparse LP algorithms by introducing sparsity constraints into the LP framework. Sparse LP has been shown to be effective in several issues related to speech modeling and coding. However, all existing approaches assume the speech signal to be minimum-phase. Because speech is known to be mixed-phase, the resulting residual signal contains a persistent maximum-phase component. The aim of this paper is to propose a novel technique which incorporates a modeling of the maximum-phase contribution of speech, and can be applied to any filter representation. The proposed method is shown to significantly increase the sparsity of the LP residual signal and to be effective in two illustrative applications: speech polarity detection and excitation modeling.
Code (0)
등록된 구현이 없습니다.
Tasks
PredictionSimilar Papers 제목 키워드 기반
Maximum Entropy Models from Phase Harmonic Covariances
The covariance of a stationary process $X$ is diagonalized by a Fourier transform. It does not take into account the complex Fourier phase and defines Gaussian maximum entropy models. We introduce a general family of pha…
Prediction of daily maximum ozone levels using Lasso sparse modeling method
This paper applies modern statistical methods in the prediction of the next-day maximum ozone concentration, as well as the maximum 8-hour-mean ozone concentration of the next day. The model uses a large number of candid…
Dimensionality ReductionPredictionData-Driven Prediction of Dynamic Interactions Between Robot Appendage and Granular Material
An alternative data-driven modeling approach has been proposed and employed to gain fundamental insights into robot motion interaction with granular terrain at certain length scales. The approach is based on an integrati…
Dimensionality ReductionRobot NavigationTight bounds for maximum $\ell_1$-margin classifiers
Popular iterative algorithms such as boosting methods and coordinate descent on linear models converge to the maximum $\ell_1$-margin classifier, a.k.a. sparse hard-margin SVM, in high dimensional regimes where the data …
Long-time predictive modeling of nonlinear dynamical systems using neural networks
We study the use of feedforward neural networks (FNN) to develop models of nonlinear dynamical systems from data. Emphasis is placed on predictions at long times, with limited data availability. Inspired by global stabil…
Data Augmentation