Patent
Speech processing system and method using wavelet packet transform
Jung Sung Il,Kwon Young Hun,Yang Sung Il +2 more
- 11 Dec 2006
6
TL;DR: In this article, a speech processing system using a wavelet packet transform and a method thereof are provided to exactly estimate components for estimating the noise from a variable signal, which is called adaptive noise estimation.
read more
Abstract: A speech processing system using a wavelet packet transform and a method thereof are provided to exactly estimate components for estimating the noise from a variable signal. A wavelet transformation member(110) transforms a speech signal contaminated by the noise and produces a transformation signal. A recursive analyzing member(120) analyzes the size of the transformation signal, that is, a transformation coefficient produced from the wavelet transformation member. An adaptive noise estimation member(130) calculates an estimation noise of the present frame from an adaptive forgetting component. A speech processing member(140) subtracts the estimation noise from the speech signal so that the noise is removed.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
Patent
Procedure for processing noisy speech signals, and apparatus and computer program therefor
Sung Il Jung,Dong Gyung Ha +1 more
- 27 Nov 2008
TL;DR: In this paper, a noisy speech signal processing method is presented, which includes calculating a transformed spectrum by transforming an input noisy speech signals to a frequency domain, calculating a smoothed magnitude spectrum by reducing magnitude differences of the transformed spectrum between neighboring frames, calculating an estimated noise component of the smoothed spectrum, and calculating an identification ratio which represents a ratio of a noise component included in the input noisy signals, by using the smootheme spectrum and the search spectrum.
9
Patent
Method for processing noisy speech signal, apparatus for same and computer-readable recording medium
Sung Il Jung,Dong Gyung Ha +1 more
- 31 Mar 2009
TL;DR: In this article, a noise estimation method for a noisy speech signal according to an embodiment of the present invention includes the steps of approximating a transformation spectrum by transforming an input noise signal to a frequency domain, calculating a smoothed magnitude spectrum having a decreased difference in a magnitude of the transformation spectrum between neighboring frames and calculating a search spectrum to represent an estimated noise component of the smoothed spectrum, and estimating a noise spectrum by using a recursive average method using an adaptive forgetting factor defined by using the search spectrum.
5
Patent
A signal processing method for analyzing combustion noise based on multiple injection conditions
Liu Long,Chen Ximing,Jingtao Du +2 more
- 11 Dec 2018
TL;DR: In this article, a signal processing method for analyzing combustion noise based on multiple injection conditions was proposed, which decomposes a cylinder pressure signal of a diesel engine under the condition of secondary injection by a three-layer stationary wavelet packet decomposition.
2
Patent
Speech emotion feature extraction method based on long-time-frame background noise estimation
Huang Yongming,Tian Kexin,Ma Ziying +2 more
- 03 Aug 2018
TL;DR: In this paper, a speech emotion feature extraction method utilizes asymmetric noise suppression to perform noise compensation on a speech signal in a frequency band of wavelet packet basis decomposition, and finally performs spectral centroid weighting on feature energy, and further increases noise robustness of extracted features.
2
Speech Enhancement by Overweighting Gain with Nonlinear Structure in Wavelet Packet Transform
TL;DR: A sound quality enhancement method by overweighting gain of a nonlinear structure in a wavelet packet area is provided to restrain the generation of musical noise efficiently and ensure reliable intelligibility in an enhanced voice.
2
References
Patent
Voice activity detection method and device
Joachim Stegmann,Gerhard Dipl.-Ing. Schröder +1 more
- 22 Apr 1998
TL;DR: In this article, a wavelet transformation is computed for each frame from which a set of parameters is determined, from which the decision variables control a decision logic circuit, whose result after smoothing in a time smoothing circuit, provides the statement “speech present/no speech”.
15
Patent
Voice activity detection
Adil Benyassine,Eyal Shlomot +1 more
- 20 Jan 1997
TL;DR: In this article, a method for generating frame voicing decisions for an incoming speech signal having periods of active voice and non-active voice for a speech encoder 110 in a speech communication system is described.
10
Patent
Feature quantity extracting method for automatic voice recognition
Luca Rigazio,Critze David,Applebaum Ted,Jean-Claude Junqua +3 more
- 06 Jul 2001
TL;DR: In this paper, a voice signal is decomposed into wavelets arrayed in asymmetrical tree data constitution, and local features needed to model a sound of a specific class are extracted and used to optimize sounds of various classes by the single recognition device.
5
Patent
Device and a method for removing an audio noise by using a wavelet transform function, particularly with regards to removing the audio noise while outputting audio data as a high-quality audio sound
Kang Jin Su
- 29 Jan 2005
TL;DR: In this paper, a wavelet transform algorithm unit was proposed to remove a noise included in the audio data of the ASCII code by performing a Wavelet transform process, thereby outputting audio data as a high-quality audio sound at the request of a user.
1
Related Papers (5)
N.S.A. Kadel,A.M. Refat +1 more
- 23 Feb 1999
Yuriy Romanyshyn,Viktor Tkachenko +1 more
- 07 Apr 2011
Jizeng Wang,Chanfei Wang +1 more
- 27 May 2008