A method for statistical parametric speech synthesis incorporating generative adversarial networks (GANs) is proposed. Although powerful deep neural networks (DNNs) techniques can be applied to artificially synthesize speech waveform, the synthetic speech quality is low compared with that of natural speech. One of the issues causing the quality degradation is an over-smoothing effect often observed in the generated speech parameters. A GAN introduced in this paper consists of two neural networks: a discriminator to distinguish natural and generated samples, and a generator to deceive the discriminator. In the proposed framework incorporating the GANs, the discriminator is trained to distinguish natural and generated speech parameters, while the acoustic models are trained to minimize the weighted sum of the conventional minimum generation loss and an adversarial loss for deceiving the discriminator. Since the objective of the GANs is to minimize the divergence (i.e., distribution difference) between the natural and generated speech parameters, the proposed method effectively alleviates the over-smoothing effect on the generated speech parameters. We evaluated the effectiveness for text-to-speech and voice conversion, and found that the proposed method can generate more natural spectral parameters and F0 than conventional minimum generation error training algorithm regardless its hyper-parameter settings. Furthermore, we investigated the effect of the divergence of various GANs, and found that a Wasserstein GAN minimizing the Earth-Mover's distance works the best in terms of improving synthetic speech quality.Index Terms-Statistical parametric speech synthesis, textto-speech synthesis, voice conversion, deep neural networks, generative adversarial networks, over-smoothing.

No abstract

This paper presents a deep neural network (DNN)-based phase reconstruction from amplitude spectrograms. In audio signal and speech processing, the amplitude spectrogram is often used for processing, and the corresponding phase spectrogram is reconstructed from the amplitude spectrogram on the basis of the Griffin-Lim method. However, the Griffin-Lim method causes unnatural artifacts in synthetic speech. Addressing this problem, we introduce the von-Mises-distribution DNN for phase reconstruction. The DNN is a generative model having the von Mises distribution that can model distributions of a periodic variable such as a phase, and the model parameters of the DNN are estimated on the basis of the maximum likelihood criterion. Furthermore, we propose a group-delay loss for DNN training to make the predicted group delay close to a natural group delay. The experimental results demonstrate that 1) the trained DNN can predict group delay accurately more than phases themselves, and 2) our phase reconstruction methods achieve better speech quality than the conventional Griffin-Lim method.

Voice conversion (VC) using sequence-to-sequence learning of context posterior probabilities is proposed. Conventional VC using shared context posterior probabilities predicts target speech parameters from the context posterior probabilities estimated from the source speech parameters. Although conventional VC can be built from non-parallel data, it is difficult to convert speaker individuality such as phonetic property and speaking rate contained in the posterior probabilities because the source posterior probabilities are directly used for predicting target speech parameters. In this work, we assume that the training data partly include parallel speech data and propose sequence-to-sequence learning between the source and target posterior probabilities. The conversion models perform non-linear and variable-length transformation from the source probability sequence to the target one. Further, we propose a joint training algorithm for the modules. In contrast to conventional VC, which separately trains the speech recognition that estimates posterior probabilities and the speech synthesis that predicts target speech parameters, our proposed method jointly trains these modules along with the proposed probability conversion modules. Experimental results demonstrate that our approach outperforms the conventional VC.

We developed a novel cationic antimicrobial peptide, AG30/5C, which demonstrates angiogenic properties similar to those of LL-37 or PR39. However, improvement of its stability and cost efficacy are required for clinical application. Therefore, we examined the metabolites of AG30/5C, which provided the further optimized compound, SR-0379. SR-0379 enhanced the proliferation of human dermal fibroblast cells (NHDFs) via the PI3 kinase-Akt-mTOR pathway through integrin-mediated interactions. Furthermore SR-0379 promoted the tube formation of human umbilical vein endothelial cells (HUVECs) in co-culture with NHDFs. This compound also displays antimicrobial activities against a number of bacteria, including drug-resistant microbes and fungi. We evaluated the effect of SR-0379 in two different would-healing models in rats, the full-thickness defects under a diabetic condition and an acutely infected wound with full-thickness defects and inoculation with Staphylococcus aureus. Treatment with SR-0379 significantly accelerated wound healing when compared to fibroblast growth factor 2 (FGF2). The beneficial effects of SR-0379 on wound healing can be explained by enhanced angiogenesis, granulation tissue formation, proliferation of endothelial cells and fibroblasts and antimicrobial activity. These results indicate that SR-0379 may have the potential for drug development in wound repair, even under especially critical colonization conditions.

No abstract

No abstract

SUMMARY(1) Records of 52 patients with aortitis syndrome were reviewed. The male to female ratio was 1: 4.2, and the age of the onset averaged 23.6-year-old.(2) There was a tendency that the lesions of the thoracic and abdominal aorta types involved younger patients than the aortic arch type. The observation of clinical course suggested the distribution of arterial lesions was fixed in the initial period of the disease.(3) It was speculated that preceding infections, especially nasopharyngeal and pulmonary infections, may facilitate an autoimmune process which has been considered to be related to the development of aortitis syndrome.(4) The principal arterial changes were stenosis and occlusion, but dilatation or aneurysm formation was found in some cases. The distribution of these lesions was analysed.(5) The symptoms were classified into the principal symptoms and the accompanying symptoms, and their time relationship was reviewed.(6) Emphasis was made on the importance of careful palpation of bilateral radial pulses, blood pressure measurement in the both arms and legs and careful auscultation of cardiac and vascular murmurs in the diagnosis of aortitis syndrome. Frequent occurrence of renovascular hypertension and aortic insufficiency associated with aortitis syndrome was pointed out. Ischemic retinal changes were less frequently found than hypertensive changes.(7) Laboratory data were compared between the cases of active inflammatory stage and those of inactive stage.(8) Early effects of steriod hormones were evaluated.