一种基于小波和快速傅里叶变换的学习型歌唱系统

语音识别和合成技术分别实现了计算机理解人类语言和模仿人类阅读文本的功能,提出了一种实现计算机学习并演唱歌曲的系统。系统运用敲击定位法定位发音时刻,然后利用Daubechies小波变换和快速傅里叶变换计算出对应的基频,采用语音合成技术输出声音。     

Watermarking via zero assigned filter banks

In order to identify the owner and distributor of digital data, a watermarking scheme in frequency domain for multimedia files is proposed. The scheme satisfies the imperceptibility and persistence requirements and it is robust against additive noise. It consists of a few stages of wavelet decomposition of several subblocks of the original signal using special zero assigned filter banks. By assigning zeros to filters on the high frequency portion of the spectrum, filter banks with frequency selective response are obtained. The information is then inserted in the wavelet-decomposed and compressed signal. Several robustness tests are performed on male voice, female voice, and music files, color and gray level images. The algorithm is tested under white Gaussian noise and against JPEG compression and it is observed to be robust even when exposed to high levels of corruption.     

采用小波变换的立体匹配:一种基于相位的方法

本义提出了一种采用小波变换基于相位的立体匹配算法。该算法引入多分辨率分析思想,借助小波变换的Ｍａｌｌａｔ算法,采用金字塔式的多尺度匹配结构,从而显著地提高了匹配效率。同时,该算法利用平滑性约束条件实现了相位不稳定区域视差的自适应插值,保证了良好的匹配效果。立体象对的测试结果表明了该算法的有效性。     

小波波形收敛的充要条件

本义在文献[４]的基础上证明了小波滤波器波形收敛的充分条件,完善了小波波形收敛的充要条件。     

A taxonomy for wavelet neural networks applied to nonlinear modelling

This article presents a novel classification of wavelet neural networks based on the orthogonality/non-orthogonality of neurons and the type of nonlinearity employed. On the basis of this classification different network types are studied and their characteristics illustrated by means of simple one-dimensional nonlinear examples. For multidimensional problems, which are affected by the curse of dimensionality, the idea of spherical wavelet functions is considered. The behaviour of these networks is also studied for modelling of a low-dimension map.     

Wavelet based non-parametric NARX models for nonlinear input–output system identification

Wavelet based non-parametric additive NARX models are proposed for nonlinear input–output system identification. By expanding each functional component of the non-parametric NARX model into wavelet multiresolution expansions, the non-parametric estimation problem becomes a linear-in-the-parameters problem, and least-squares-based methods such as the orthogonal forward regression (OFR) approach, coupled with model size determination criteria, can be used to select the model terms and estimate the parameters. Wavelet based additive models, combined with model order determination and variable selection approaches, are capable of handling problems of high dimensionality.     

Structural damage detection with statistical analysis from support excitation

The unit impulse response (UIR) functions obtained from a structure under support excitation are used to identify local structural damages. The extraction of the UIR with the aid of discrete wavelet transform from the measured acceleration is described. The sensitivity matrix of these UIRs is then obtained based on the finite element model and the time-stepping integration method. Based on the computed sensitivity matrix of the UIRs from several accelerometers, a two-step model updating method is adopted for identifying the local damages. Statistical analysis is included into the damage identification procedure with the measurement noise taken as an independent random variable in the measured UIRs. A new damage localization index is proposed and the mean values of the identified parameters are taken as the damage severity. Finally, a nine-bay three-dimensional frame structure is analyzed numerically and experimentally using the proposed technique. The damage scenarios of multiple damages with different levels of noise are considered. The identified results are shown satisfactory, indicating the feasibility and effectiveness of the proposed technique.     

Reduction of transients during lifting based spatial switching of two-channel filter banks

Time/space varying filter banks (FBs) are useful for non-stationary images. Lifting factorization of FBs results in structural perfect reconstruction even during the transition from one FB to other. This allows spatial switching between arbitrary FBs, avoiding the need to design border FBs. However, we show that lifting based switching between arbitrarily designed FBs induces spurious transients in the subbands during the transition. In this paper, we study the transients in lifting based switching of two-channel FBs. We propose two solutions to overcome the transients. One solution consists of a boundary handling mechanism to switch between any arbitrarily designed FBs, while the other solution proposes to design the FBs with a set of conditions applied on lifting steps. Both solutions maintain good frequency response during the transition and eliminate the transients. Using the proposed methods, we develop a spatial adaptive transform by switching between the long length FBs (either the JPEG2000 9/7 FB or the newly designed 13/11 FB) and the short length FBs (JPEG2000 5/3 FB) for lossy image compression. This adaptive transform shows PSNR improvement for images over JPEG2000 9/7 FB in low bit rate region (up to 0.2 bpp) and subjective improvements with reduced ringing up to medium bit rates (up to 0.6 bpp).     

The wavelet-based theory of spatial naturally curved and twisted linear beams

The paper presents the wavelet-based discretization of the linearized finite-strain beam theory which assumes small displacements, rotations and strains but is capable of considering an arbitrary initial geometry and material behaviour. In the numerical solution algorithm, we base our derivations on the vector of strain measures as the only unknown functions in a finite element. In such a way the determination of the beam quantities does not require the differentiation. This is an important advantage which allows a wider range of shape functions. In the present paper, the classical polynomial interpolation is compared to scaling and wavelet function interpolations. The computational efficiency of the method is demonstrated by analyzing initially curved and twisted beams.     

Nonlinear system identification using Wiener type Laguerre-Wavelet network model

Compact nonlinear black box models, capable of representing highly nonlinear systems, are of high demand both in industry and academia. In this paper it has been shown that the use of Laguerre basis filters coupled with a wavelet network in Wiener type model structure are capable of modeling highly nonlinear systems with acceptable accuracy. Although individual merits of orthonormal basis functions (especially Laguerre filters) and multiresolution wavelet decompositions and/or wavelet network have been well documented in various literatures on system identification in the past, their combinational advantages in nonlinear system identification has never been explored. Laguerre filter models have the ability to approximate linear systems (even with time delay) with a model order lower than the traditional ARX (e.g., FIR, AR) modeling. Use of Laguerre models for mildly nonlinear system is possible only with a piece-wise linear models. Wavelet basis functions have the property of localization in both time and frequency and they can approximate even severe nonlinearities with appreciable accuracy with fewer model terms. But wavelet approximations fail miserably, in terms of model parsimony, if used for approximating linear or mildly nonlinear systems. In the present work a Laguerre-Wavelet network model has been proposed which combines the Laguerre and Wavelet approximations. In the said model, merits of both these approximations are retained whereas their demerits are suppressed.