Segregation of voiced and unvoiced components from residual of speech signal

In conventional source-filter models, voiced and unvoiced components were considered independently. However, in practice it was difficult to separate the source into two parts. An actual source consists of a mixture of two sources and the ratio varies according to the content or the intention of speaker. It had been investigated to separate the voiced and unvoiced components for different source models. Source signals were modeled based on the residual signal measured from inverse filtering. Three different source models were assumed. The parameters of each model were optimized for the original speech signal using a genetic algorithm. The resulting parameters were compared in terms of the mel-cepstral distance to the original signal, the spectrogram and the spectral envelope from the synthesized signal. The optimization method achieves an improvement of 15% for the Klatt model, but there is little improvement in the modified residual case.

Seismic data denoising based on learning-type overcomplete dictionaries

The transform base function method is one of the most commonly used techniques for seismic denoising, which achieves the purpose of removing noise by utilizing the sparseness and separateness of seismic data in the transform base function domain. However, the effect is not satisfactory because it needs to pre-select a set of fixed transform-base functions and process the corresponding transform. In order to find a new approach, we introduce learning-type overcomplete dictionaries, i.e., optimally sparse data representation is achieved through learning and training driven by seismic modeling data, instead of using a single set of fixed transform bases. In this paper, we combine dictionary learning with total variation （TV） minimization to suppress pseudo-Gibbs artifacts and describe the effects of non-uniform dictionary sub-block scale on removing noises. Taking the discrete cosine transform and random noise as an example, we made comparisons between a single transform base, non-learning-type, overcomplete dictionary and a learning-type overcomplete dictionary and also compare the results with uniform and nonuniform size dictionary atoms. The results show that, when seismic data is represented sparsely using the learning-type overcomplete dictionary, noise is also removed and visibility and signal to noise ratio is markedly increased. We also compare the results with uniform and nonuniform size dictionary atoms, which demonstrate that a nonuniform dictionary atom is more suitable for seismic denoising.

AUDIO STEGANALYSIS OF DSSS BASED ON STATISTICAL MOMENTS OF HISTOGRAM

Compared with the histogram of Discrete Cosine Transform (DCT) coefficients before the Direct Sequence Spread Spectrum (DSSS) embedding, the peak value of the histogram after the embedding decreases and expands toward the border. Based on the property, an audio steganalysis of DSSS based on statistical moments of histogram is proposed. The statistical moments of the histogram in DCT domain and its frequency domain and the statistical moments of the histogram of the wavelet coefficients of every level in frequency domain are calculated as the features of classification. Support Vector Machine (SVM) is exploited as the classifier. Experimental results show that the proposed technique is effective on the DSSS embedding in DCT domain using different embedding length, and the average detection rate is 91.75%.

余音

阳光透过稀疏的枝叶像过筛一样在地面留下树影斑驳。尘封的戏台再次开启，那段美丽悠远的芗剧，一唱便是几千几百个花季雨季。记忆中，每当有戏班来时，爷爷总爱抱上年幼的我前去观看。芗剧的戏台虽然简陋但却满是古色古香，不用反绒红缎的大幕，只需几张镂花的木桌木椅也不乏一股韵味。

Musical Isomorphisms and Problems of Lifts

Using the complete lift on tangent bundles, the authors construct the complete lift on cotangent bundles of tensor fields with the aid of a musical isomorphism. In this new framework, the authors have a new intrepretation of the complete lift of tensor fields on cotangent bundles.