Non-parallel voice conversion (VC) is a technique for learning the mapping from source to target speech without relying on parallel data. This is an important task, but it has been challenging due to ...the disadvantages of the training conditions. Recently, CycleGAN-VC has provided a breakthrough and performed comparably to a parallel VC method without relying on any extra data, modules, or time alignment procedures. However, there is still a large gap between the real target and converted speech, and bridging this gap remains a challenge. To reduce the gap, we propose CycleGAN-VC2, which is an improved version of CycleGAN-VC incorporating three new techniques: an improved objective (two-step adversarial losses), improved generator (2-1-2D CNN), and improved discriminator (PatchGAN). We evaluated our method on a non-parallel VC task and analyzed the effect of each technique in detail. An objective evaluation showed that these techniques help bring the converted feature sequence closer to the target in terms of both global and local structures, which we assess by using Mel-cepstral distortion and modulation spectra distance, respectively. A subjective evaluation showed that CycleGAN-VC2 outperforms CycleGAN-VC in terms of naturalness and similarity for every speaker pair, including intra-gender and inter-gender pairs. 1
This research explores the mechanical properties and erosion mechanisms of SS316, VC+TiC, and PTFE coatings to study cavitation erosion resistance. Microhardness analysis revealed that VC+TiC showed ...the highest resilience due to its hardness, while PTFE, despite lower hardness, displayed potential with formulation adjustments. Lasso Regression models effectively predicted mass loss under various testing conditions, indicating strong relationships between independent variables and erosion responses. Scanning Electron Microscopy (SEM) analysis uncovered distinct erosion mechanisms for each coating, with SS316 showing plastic deformation, VC+TiC exhibiting crater formation, and PTFE displaying torn splats. The development of superhydrophobic surfaces from VC+TiC and PTFE coatings offers promising applications in mitigating cavitation erosion, providing valuable material-specific insights into this phenomenon.
Display omitted
A hybrid solid electrolyte interphase (SEI) formation additive, vinylene carbonate (VC)–LiNO3, was investigated in carbonic ester electrolytes. An efficiency of lithium plating/stripping as high as ...nearly 100% and spherical Li deposits were obtained. The electrochemical impedance spectroscopy (EIS) results demonstrate that the modified SEI is very stable and of good conductivity. X-ray photoelectron spectroscopy (XPS) results indicate that VC–LiNO3 dominates the surface chemistry of the Li anode. The formation of Li3N in the SEI contributes to the enhancement of the anode performance.
•Vinylene carbonate (VC)–LiNO3 was employed in carbonic ester electrolytes for the first time.•One of the main components in modified SEI film by LiNO3 is Li3N which is obviously distinct from the ones generated in ether-based electrolytes.•An efficiency of Li plating/stripping near to 100% and a uniform morphology of cycled Li with spherical deposits were obtained.
Many-to-many voice conversion (VC) is a technique aimed at mapping speech features between multiple speakers during training and transferring the vocal characteristics of one source speaker to ...another target speaker, all while maintaining the content of the source speech unchanged. Existing research highlights a notable gap between the original and generated speech samples in terms of naturalness within many-to-many VC. Therefore, there is substantial room for improvement in achieving more natural-sounding speech samples for both parallel and nonparallel VC scenarios. In this study, we introduce a generative adversarial network (GAN) system with a guided loss (GLGAN-VC) designed to enhance many-to-many VC by focusing on architectural improvements and the integration of alternative loss functions. Our approach includes a pair-wise downsampling and upsampling (PDU) generator network for effective speech feature mapping (FM) in multidomain VC. In addition, we incorporate an FM loss to preserve content information and a residual connection (RC)-based discriminator network to improve learning. A guided loss (GL) function is introduced to efficiently capture differences in latent feature representations between source and target speakers, and an enhanced reconstruction loss is proposed for better contextual information preservation. We evaluate our model on various datasets, including VCC 2016, VCC 2018, VCC 2020, and an emotional speech dataset (ESD). Our results, based on both subjective and objective evaluation metrics, demonstrate that our model outperforms state-of-the-art (SOTA) many-to-many GAN-based VC models in terms of speech quality and speaker similarity in the generated speech samples.
•This study constructed a new method for the detection of vitamin C and H2PO4-.•The constructed supramolecular sensor exhibits good AIE performance.•This chemosensor was highly specific and sensitive ...to the detection of ascorbic acid and H2PO4-.•The sensor 2BCP-@Fe3+ shows an outstanding anti-interference ability for the detection of ascorbic acid and H2PO4-.
The benefits of Vc for the body's typical metabolism have received adequate attention. However, effective probes for analyzing specific Vc are currently lacking. To monitor Vc levels, we design and create a novel luminescence-silent oxidation AIE fluorescent probe (2BCP@Fe) by interacting between a coumarin dual-functionalized fluorescent pillar5arene derivative and Fe3+ to monitor largescale Vc. The morphological transformation was achieved by adding a certain amount of aqueous solution into the BCP DMF solution. A fluorescence ‘turn-on‘ phenomenon was observed, which was caused by restricting intramolecular rotation.
Meanwhile, SEM measurements supported the importance of the cation-π interaction between Fe3+ and BCP in this supramolecular process. Additionally, the supramolecular system (2BCP@Fe) can monitor H2PO4- ions through competitive contact and can operate as a “turn-on” fluorescent probe for Vc detection through the process of redox reaction. As a result, such a novel fabrication strategy might offer a prospective strategy for preparing the metal-modified macrocyclic host functional materials for sensing, catalysis, and other fields.
This work investigates the critical roles of two-step sintering (TSS) and laminated structure on the sintering behavior and mechanical properties of functionally graded WC-TiC-Al2O3 nanostructured ...composite materials doped with Cr3C2/VC. Results show that excellent mechanical properties are achieved for tailored TSS conditions with a hardness of 27.91 ± 2.3 GPa and a flexural strength of 1423.3 ± 23.5 MPa. The desirable mechanical properties are attributed to the suppressed grain growth without densification deterioration. TSS is more effective in facilitating the favorable dispersion of secondary phase toughening nano-particulates in a WC matrix than conventional sintering (CS). Cr3C2/VC dopant plays an important role in maximizing and shifting the temperature range of the kinetic window for WC-Al2O3 composites. Al2O3 crack deflection, transgranular Al2O3, microcracking, WC crack bridging and plate-like WC crack deflection are the major toughening mechanisms. Residual surface compressive stress induced by the graded structure is also an appreciated contribution to the improvement of mechanical properties.
The effect of ultrasonic intensity (28.14, 70.35, and 112.56 W/cm2) on Lignosus rhinocerotis polysaccharide (LRP) degraded by ultrasound assisted H2O2/Vc system (U-H/V) was investigated. U-H/V broke ...the molecular chain of LRP and improved the conformational flexibility, decreasing the molecular weight, intrinsic viscosity (η) and particle size. The functional groups and hyperbranched structure of LRP were almost stable after U-H/V treatment, however, the triple helix structure of LRP was partially disrupted. With increasing ultrasonic intensity, the critical aggregation concentration increased from 0.59 mg/mL to 1.57 mg/mL, and the hydrophobic microdomains reduced. Furthermore, the LRP treated with U-H/V significantly inhibited HepG2 cell proliferation by inducing apoptosis. The increase in antitumor activity of LRP was closely associated with the reduction of molecular weight, η, particle size and hydrophobic microdomains. These results revealed that U-H/V treatment facilitates the degradation of LRP and provides a better insight into the structure-antitumor activity relationship of LRP.
Display omitted
•Ultrasound-assisted H2O2/Vc system (U-H/V) exhibited favorable degradation on LRP.•U-H/V broke the molecular chain of LRP and improved the conformational flexibility.•U-H/V disrupted part of the helical structure of LRP.•Hydrophobic microdomains of LRP decreased with increasing ultrasonic intensity.•The antitumor activity of LRP was improved after U-H/V treatment.
Display omitted
•The tightly coupled of Mo-VC with CdS photocatalyst by the solid-liquid mixing method.•The Mo-VC can efficiently catalyze H2 evolution reaction than VC under visible light.•An AQE of ...4.3 % for H2 evolution is obtained on the optimized CdS/Mo-VC at 420 nm.•The catalytic H2evolution activity of Mo-VCis higher than noble metal (Pt、Au、Ag).
Seeking an efficient and stable cocatalyst is of importance for semiconductor-based photocatalytic water splitting to produce hydrogen. In this sense, Mo-activation of VC for better cocatalytic properties was proposed for the first time, aiming to attain improved hydrogen generation of CdS. Our novel, yet facile physical preparation method permits intimate coupling of Mo-activated VC cocatalyst to CdS (through electrostatic attraction force), thereby promises drastic activity enhancement for hydrogen generation. Results indicate the best VC-activating competency of Mo against other transition metals (Co, Fe, Ni, Zn and Cu), which induces highest H2 productivity of CdS after modifying with Mo-activated VC. Under optimum setting, CdS/Mo(0.38)-VC(5 wt%) manifested highest photocatalytic hydrogen production rate of 2267 μmol h−1 g−1, associated to 300 % and 440 % improvement as compared to that of CdS/VC(5 wt%) and pure CdS, respectively. Such activity also outperformed those that yielded by MoSx- or noble metals(Pt, Au or Ag)-co-catalyzed CdS, attributed to the promoted interfacial charges transfer and lowered overpotential of CdS/Mo(0.38)-VC(5 wt%) in hydrogen generation. Correspondingly, high apparent hydrogen production quantum efficiency of 4.3 % was also recorded by the same photocatalyst at 420 nm. In term of stability test, CdS/Mo(0.38)-VC(5 wt%) manifested excellent longevity in 24 h continuous photoreaction and fair recyclability over five consecutive photoreaction cycles. This work provides a new insight to cocatalysts design and synthesis, thereby realizing economic alternatives to precious metal (Pt, Au, Ag) in solar-derived H2 production.
Recently, researchers in the artificial neural network field have focused their attention on connectionist models composed by several hidden layers. In fact, experimental results and heuristic ...considerations suggest that deep architectures are more suitable than shallow ones for modern applications, facing very complex problems, e.g., vision and human language understanding. However, the actual theoretical results supporting such a claim are still few and incomplete. In this paper, we propose a new approach to study how the depth of feedforward neural networks impacts on their ability in implementing high complexity functions. First, a new measure based on topological concepts is introduced, aimed at evaluating the complexity of the function implemented by a neural network, used for classification purposes. Then, deep and shallow neural architectures with common sigmoidal activation functions are compared, by deriving upper and lower bounds on their complexity, and studying how the complexity depends on the number of hidden units and the used activation function. The obtained results seem to support the idea that deep networks actually implements functions of higher complexity, so that they are able, with the same number of resources, to address more difficult problems.
This paper provides an overview of Scalable High efficiency Video Coding (SHVC), the scalable extensions of the High Efficiency Video Coding (HEVC) standard, published in the second version of HEVC. ...In addition to the temporal scalability already provided by the first version of HEVC, SHVC further provides spatial, signal-to-noise ratio, bit depth, and color gamut scalability functionalities, as well as combinations of any of these. The SHVC architecture design enables SHVC implementations to be built using multiple repurposed single-layer HEVC codec cores, with the addition of interlayer reference picture processing modules. The general multilayer high-level syntax design common to all multilayer HEVC extensions, including SHVC, MV-HEVC, and 3D HEVC, is described. The interlayer reference picture processing modules, including texture and motion resampling and color mapping, are also described. Performance comparisons are provided for SHVC versus simulcast HEVC and versus the scalable video coding extension to H.264/advanced video coding.
PDF
