Unsupervised intra-lingual and cross-lingual speaker adaptation for HMM-based speech synthesis using two-pass decision tree construction

Hidden Markov model (HMM)-based speech synthesis systems possess several advantages over concatenative synthesis systems. One such advantage is the relative ease with which HMM-based systems are adapted to speakers not present in the training dataset. Speaker adaptation methods used in the field of HMM-based automatic speech recognition (ASR) are adopted for this task. In the case of unsupervised speaker adaptation, previous work has used a supplementary set of acoustic models to estimate the transcription of the adaptation data. This paper first presents an approach to the unsupervised speaker adaptation task for HMM-based speech synthesis models which avoids the need for such supplementary acoustic models. This is achieved by defining a mapping between HMM-based synthesis models and ASR-style models, via a two-pass decision tree construction process. Second, it is shown that this mapping also enables unsupervised adaptation of HMM-based speech synthesis models without the need to perform linguistic analysis of the estimated transcription of the adaptation data. Third, this paper demonstrates how this technique lends itself to the task of unsupervised cross-lingual adaptation of HMM-based speech synthesis models, and explains the advantages of such an approach. Finally, listener evaluations reveal that the proposed unsupervised adaptation methods deliver performance approaching that of supervised adaptation.

DT-MRI data visualisation using the dual tree Complex Wavelet transform

In this paper a novel visualisation method for diffusion tensor MRI datasets is introduced. This is based on the use of Complex Wavelets in order to produce "stripy" textures which depict the anisotropic component of the diffusion tensors. Grey-scale pixel intensity is used to show the isotropic component. This paper also discusses enhancements of the technique for 3D visualisation. © 2004 IEEE.

Effects of gamma-aminobutyric acid, progesterone and ionophore A23187 on acrosome reaction of tree shrew sperm in vitro: examination of acrosome reaction with an improved fluorescence microscopy

A number of acrosome reaction (AR) initiators have been found to be effective in inducing AR of human, laboratory and domestic animal sperm. Using an improved simple fluorescence microscopy, effects of gamma-aminobutyric acid (GABA), progesterone and ionophore A23187 on sperm AR of tree shrew, a useful animal model in biomedical research, have been investigated. Spontaneous AR in 4.92-7.53% of viable sperm was observed. Complete AR in 10.31-18.25% of viable tree shrew sperm was obviously induced by 5 mu M and 10 mu M calcium ionophore A23187, 1 mM GABA, and 5 mu M progesterone, and there were no significant differences between their abilities to initiate complete AR. No significant differences of AR percentages between 1- and 2-h treatments with A23187, progesterone and/or GABA were observed. These results suggested that the responses of tree shrew sperm to these AR initiators are similar to that of human and other mammalian sperm. (C) 1997 Elsevier Science B.V.

Ovulation in the tree shrew (Tupaia belangeri) induced by gonadotrophins

The induction of ovulation by exogenous gonadotrophins is an important approach for recovering oocytes used for studies on the reproductive biology of some mammals. In the present study, pregnant mare serum gonadotrophin (PMSG) and human chorionic gonadot

Making a shallow network deep: Conversion of a boosting classifier into a decision tree by boolean optimisation

This paper presents a novel way to speed up the evaluation time of a boosting classifier. We make a shallow (flat) network deep (hierarchical) by growing a tree from decision regions of a given boosting classifier. The tree provides many short paths for speeding up while preserving the reasonably smooth decision regions of the boosting classifier for good generalisation. For converting a boosting classifier into a decision tree, we formulate a Boolean optimization problem, which has been previously studied for circuit design but limited to a small number of binary variables. In this work, a novel optimisation method is proposed for, firstly, several tens of variables i.e. weak-learners of a boosting classifier, and then any larger number of weak-learners by using a two-stage cascade. Experiments on the synthetic and face image data sets show that the obtained tree achieves a significant speed up both over a standard boosting classifier and the Fast-exit-a previously described method for speeding-up boosting classification, at the same accuracy. The proposed method as a general meta-algorithm is also useful for a boosting cascade, where it speeds up individual stage classifiers by different gains. The proposed method is further demonstrated for fast-moving object tracking and segmentation problems. © 2011 Springer Science+Business Media, LLC.

Multiple classifier boosting and tree-structured classifiers

Visual recognition problems often involve classification of myriads of pixels, across scales, to locate objects of interest in an image or to segment images according to object classes. The requirement for high speed and accuracy makes the problems very challenging and has motivated studies on efficient classification algorithms. A novel multi-classifier boosting algorithm is proposed to tackle the multimodal problems by simultaneously clustering samples and boosting classifiers in Section 2. The method is extended into an online version for object tracking in Section 3. Section 4 presents a tree-structured classifier, called Super tree, to further speed up the classification time of a standard boosting classifier. The proposed methods are demonstrated for object detection, tracking and segmentation tasks. © 2013 Springer-Verlag Berlin Heidelberg.

Polychlorinated dibenzo-p-dioxin and dibenzofurans (PCDD/Fs), polybrominated diphenyl ethers (PBDEs), and polychlorinated biphenyls (PCBs) monitored by tree bark in an E-waste recycling area

To investigate the environmental levels and profiles of polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs), polybrominated diphenyl ethers (PBDEs), and polychlorinated biphenyls (PCBs), tree bark samples (n = 22) were collected from Luqiao, an E-waste recycling area, in east China in July 11-13, 2006. The average concentrations of PCDD/Fs, PBDEs, and PCBs determined by isotope dilution-high resolution gas chromatography (HRGC) coupled with high resolution mass spectrometer (HRMS) were 0.1 +/- 0.0, 1.4 +/- 0.2, and 6.5 +/- 0.8 lg g (1) lipid weight, respectively. PCDD/F-toxic equivalent (TEQ, WHO-1998), PCB-TEQs, and total dioxin-like TEQs were 1.3 +/- 0.1, 0.5 +/- 0.0, and 1.8 +/- 0.2 ng g (1) lipid weight, respectively. The profiles of these pollutants in the tree bark were also discussed. Tetra-CDFs, deca-BDE and tri-CBs were the main homologues and accounted for 47% of total PCDD/Fs, 79.3% of total PBDEs, and 33.2% of total PCBs, respectively; As for TEQs, 2,3,4,7,8-PeCDF and PCB126 were the main contributors and accounted for 36% of the total PCDD/F-TEQs and 81.2% of the total PCB-TEQs, respectively. High accumulation of PCDD/Fs, PBDEs, and PCBs detected in the tree bark indicated heavy contaminations of these pollutants in Luqiao area. (C) 2008 Elsevier Ltd. All rights reserved.

Dense active appearance models using a bounded diameter minimum spanning tree

We present a method for producing dense Active Appearance Models (AAMs), suitable for video-realistic synthesis. To this end we estimate a joint alignment of all training images using a set of pairwise registrations and ensure that these pairwise registrations are only calculated between similar images. This is achieved by defining a graph on the image set whose edge weights correspond to registration errors and computing a bounded diameter minimum spanning tree (BDMST). Dense optical flow is used to compute pairwise registration and we introduce a flow refinement method to align small scale texture. Once registration between training images has been established we propose a method to add vertices to the AAM in a way that minimises error between the observed flow fields and a flow field interpolated between the AAM mesh points. We demonstrate a significant improvement in model compactness using the proposed method and show it dealing with cases that are problematic for current state-of-the-art approaches.