Search

Since 2020, synchrotron radiation facilities in several Asia-Pacific countries have been collaborating in a major project called "SYNAPSE"(Synchrotrons for Neuroscience: an Asia-Pacific Scientific Enterprise). They use x-ray imaging to attack in a coordinated fashion one of the major issues in modern science: the structure of animal and human brains, including neurons and connections. The objective is to develop Google-like maps also including detailed structural and functional information for selected regions of interest. The sheer mass of data needed for the objective poses huge problems for the acquisition, processing, storage and use of images. In order to complete the task within a reasonable time, the key element of the SYNAPSE strategy is the parallel and coordinated work of several facilities on the same specimens. This article reviews different aspects of the enterprise, including the foundations of synchrotron radiation, coherence and of its role in advanced imaging, electron accelerators, x-ray optics and detectors. This will provide the foundation for an extensive presentation of the different components of SYNAPSE, with an overview of results already obtained within the consortium.

A numerical study predicts that a single metamolecule with magnetism and chirality has giant magnetochiral (MCh) effects at microwave frequencies. The magnetism is provided by the ferromagnetic resonance of ferrite under dc bias magnetic fields, while the chirality is provided by the spiral arrangement of dielectric cubes with Mie resonance. The dielectric and magnetic resonances interfere in the metamolecule, resulting in a two-order of magnitude enhancement of the MCh effect compared with that reported in previous studies. This prediction is verified experimentally. A unity-order directional difference in the refractive index caused by the MCh effect is also demonstrated. This study is a significant milestone in the practical use of the MCh effect.

The development of neural vocoders (NVs) has resulted in the high-quality and fast generation of waveforms. However, conventional NVs target a single sampling rate and require re-training when applied to different sampling rates. A suitable sampling rate varies from application to application due to the trade-off between speech quality and generation speed. In this study, we propose a method to handle multiple sampling rates in a single NV, called the MSR-NV. By generating waveforms step-by-step starting from a low sampling rate, MSR-NV can efficiently learn the characteristics of each frequency band and synthesize high-quality speech at multiple sampling rates. It can be regarded as an extension of the previously proposed NVs, and in this study, we extend the structure of Parallel WaveGAN (PWG). Experimental evaluation results demonstrate that the proposed method achieves remarkably higher subjective quality than the original PWG trained separately at 16, 24, and 48 kHz, without increasing the inference time. We also show that MSR-NV can leverage speech with lower sampling rates to further improve the quality of the synthetic speech., 6 pages including supplement, 3 figures, accepted for INTERSPEECH 2022. Audio samples: https://rinnakk.github.io/research/publications/MSR-NV/

The local electronic state of the N,N,N',N'-tetraglycidyl-4,4′-diaminodiphenylmethane epoxy resin cured with 4,4′-diaminodiphenylsulfone (TGDDM-DDS), one of model adhesive materials, was investigated by the O K-edge X-ray absorption spectroscopy (XAS) and X-ray emission spectroscopy (XES). The total-electron-yield XAS spectra for TGDDM-DDS as a function of the X-ray dose exhibit the exponential-like decrease in the curing-derived pre-edge intensity due to the radiation damage. The quantitative analysis on the pre-edge intensity with respect to the X-ray dose enables the determination of the critical absorbed dose for the chemical change on TGDDM-DDS as 3.83 ± 0.01 MGy at 300 K. By using resonant XES with the well-focused X-ray beam, we found that (i) the charged-particle equilibrium exists for the soft X-rays in the range of at least 0.001–0.847 mm2 beam spot and that (ii) the half of the critical absorbed dose can be regarded as the limit indicator for the non-destructive chemical-state characterization on resinous adhesive materials and interfaces using soft X-rays.

This paper proposes a hierarchical generative model with a multi-grained latent variable to synthesize expressive speech. In recent years, fine-grained latent variables are introduced into the text-to-speech synthesis that enable the fine control of the prosody and speaking styles of synthesized speech. However, the naturalness of speech degrades when these latent variables are obtained by sampling from the standard Gaussian prior. To solve this problem, we propose a novel framework for modeling the fine-grained latent variables, considering the dependence on an input text, a hierarchical linguistic structure, and a temporal structure of latent variables. This framework consists of a multi-grained variational autoencoder, a conditional prior, and a multi-level auto-regressive latent converter to obtain the different time-resolution latent variables and sample the finer-level latent variables from the coarser-level ones by taking into account the input text. Experimental results indicate an appropriate method of sampling fine-grained latent variables without the reference signal at the synthesis stage. Our proposed framework also provides the controllability of speaking style in an entire utterance., 5 pages, accepted to INTERSPEECH 2020, demo page: https://www.rinna.jp/research/interspeech2020/

The distribution of topological charges on X-ray vortices was measured by differential Fourier space filtering microscope, differential radial Hilbert transform microscope. It was experimentally verified for the first time using a Spiral Fresnel zone plate objective lens. This X-ray microscope is highly sensitive to X-ray topological defects, such as edges and vortices, at the exit-face wave field of objects. Its efficient use is also discussed.

We investigate the metal to nonmagnetic insulator (MI) transition of MnP-type Ru pnictide RuP using hard x-ray and ultraviolet photoemission spectroscopies. The spectral weight at ${E}_{\mathrm{F}}$ is suppressed below the MI transition temperature ${T}_{\mathrm{MI}}$, while there is no appreciable change across the pseudogap temperature ${T}_{\mathrm{PG}}$. The estimated energy scale of the gap is $\ensuremath{\sim}110$ meV, which is in good correspondence to the spin gap opening observed in the previous NMR study. According to the band structure calculation, the density of states at ${E}_{\mathrm{F}}$ mainly originates from narrow bands of the Ru $4{d}_{xy}$ orbitals. Our results suggest that the fourfold-degenerate Ru $4{d}_{xy}$ orbitals are deeply related to the MI transition. Based on the photoemission results and the band structure calculation, we argue the possible origin of the MI transition for polycrystalline RuP.

We report a search for photon-photon elastic scattering in vacuum in the X-ray region at an energy in the center of mass system of \omega_{cms} =6.5 keV for which the QED cross section is \sigma_{QED} =2.5 \times 10^{-47} m^2. An X-ray beam provided by the SACLA X-ray Free Electron Laser is split and the two beamlets are made to collide at right angle, with a total integrated luminosity of (1.24 \pm 0.08) \times 10^{28} m^{-2}. No signal X rays from the elastic scattering that satisfy the correlation between energy and scattering angle were detected. We obtain a 95% C.L. upper limit for the scattering cross section of 1.9 \times 10^{-27} m^2 at \omega_{cms}=6.5 keV. The upper limit is the lowest upper limit obtained so far by keV experiments., Comment: 11 pages, 2 Postscript figures, 1 tables

This chapter overviews magnetochiral (MCh) effects for the X-band microwaves by a single metamolecule consisting of a copper chiral structure and a ferrite rod. The directional birefringence due to the MCh effects is induced at the resonant optical activity frequencies by applying a weak DC magnetic field of 1 mT and increased with the magnetic field. The nonreciprocal differences in refractive indices by the MCh effects are evaluated to be \(10^{-3}\) at 200 mT, which is much larger than that observed in natural chiral molecules at the visible frequencies. Moreover, the enhanced MCh effects can be obtained at ferromagnetic resonance frequencies by the ferrite rod in the metamolecule. The present study paves the way toward the realization of synthetic gauge fields for electromagnetic waves and the emergence of meta material-science using microwave metamaterials. Furthermore, higher frequencies including the visible region are accessible by our concept, in which an interaction between magnetism and chirality in the metamaterials is realized without intrinsic electronic interactions.

This paper proposes a method of selecting training data for many-to-one singing voice conversion (VC) from waveform data on the social media music app “nana.” On this social media app, users can share sounds such as speaking, singing, and instrumental music recorded by their smartphones. The number of hours of accumulated waveform data has exceeded one million, and it is regarded as “big data.” It is widely known that big data can create huge values by advanced deep learning technology. A lot of post data of multiple users having sung the same song is contained in nana's database. This data is considered suitable training data for VC. This is because VC frameworks based on statistical approaches often require parallel data sets that consist of pairs of waveform data of source and target singers who sing the same phrases. The proposed method can compose parallel data sets that can be used for many-to-one statistical VCs from nana's database by extracting frames that have small differences in the timing of utterances, based on the results of dynamic programming (DP) matching. Experimental results indicate that a system that uses training data composed by our method can convert acoustic features more accurately than a system that does not use the method.

An image recognition method based on convolutional neural networks (CNNs) using features generated from separable lattice hidden Markov models (SLHMMs) is proposed. A major problem in image recognition is that the recognition performance is degraded by geometric variations in the size and position of the object to be recognized. To solve this problem, SLHMMs have been proposed as an extension of HMMs with size and locational invariances based on state transitions. Although SLHMMs are generative models that can represent the generation processes of observations well, there is a possibility that they are not specialized for discrimination compared to discriminative models. Our method integrates SLHMMs that extract features invariant to geometric variations with CNNs that build an accurate classifier based on discriminative models with the extracted features. Face recognition experiments showed that the proposed method improves recognition performance.

This paper proposes an image recognition method based on separable lattice hidden Markov models (SLHMMs) using a deep neural network (DNN) for output probability distributions. The geometric variations of the object to be recognized, e.g., size and location, are essential in image recognition. SLHMMs, which have been proposed to reduce the effect of geometric variations, can perform elastic matching both horizontally and vertically. Gaussian distributions are typical for modeling the output distribution of SLHMMs. However, these distributions may not be sufficient to represent patterns of image regions. Our method integrates SLHMMs and a DNN and can be used to model an image effectively by explicit modeling of the generative process based on SLHMMs and advanced feature classification based on a DNN. image recognition experiments showed that the proposed method improves recognition performance.

Obstructive sleep apnea syndrome (OSAS) is often associated with metabolic disorders such as obesity and type 2 diabetes and may contribute to cardiovascular events. A novel class of antidiabetic drugs, the sodium glucose cotransporter 2 inhibitors (SGLT2i) reduce body weight (BW), although there is limited data on their impact on OSAS. We therefore evaluated the effect of SGLT2i on OSAS in patients with type 2 diabetes. The presented study was a retrospective design in 18 patients with type 2 diabetes with OSAS (4 males, age range 39-81 yr) administrated a SGLT2i. HbA1c, BW, body mass index (BMI), blood pressure (BP) and apnea hypopnea index (AHI) were evaluated before and after SGLT2i administration. The relationships between the reduction in AHI and the other variables were examined using Pearson correlation analysis. We have got result that SGLT2i reduced AHI from 31.9 ± 18.0 to 18.8 ± 11.5 events per hr (p = 0.003). HbA1c, BW and BMI decreased significantly, whereas BP did not. The Pearson correlation analysis showed a significant relationship between the reduction in AHI and pre-administration of AHI. In conclusion, SGLT2i reduced not only HbA1c, BW and BMI but also AHI significantly and therefore has potential as an effective treatment of OSAS.

This paper describes the Blizzard Machine Learning Challenge (BMLC) 2017, which is a spin-off of the Blizzard Challenge. The annual Blizzard Challenges 2005–2017 were held to better understand and compare research techniques in building corpus-based text-to-speech (TTS) systems on the same data. The series of Blizzard Challenges has helped us measure progress in TTS technology. However, to get competitive performance, a lot time has to be spent on skilled tasks. This may make the Blizzard Challenge unattractive to machine learning researchers from other fields. Therefore, we recommend that the BMLC not involve these speech-specific tasks and that it allow participants to concentrate on the acoustic modeling task, framed as a straightforward machine learning problem, with a fixed dataset. In the BMLC 2017, two types of datasets consisting of four hours of speech data suitable for machine learning problems were distributed. This paper summarizes the purpose, design, and whole process of the challenge and its results.