α-MoO3 nanobelts were successfully prepared by a facile hydrothermal method with sodium molybdate (Na2MoO4) as the Mo source and NaCl as the capping agent. The as-prepared products were characterized using Fourier transformation infrared spectrophotometry (FT-IR), X-ray powder diffraction (XRD), field emission scanning electron microscopy (FESEM), transmission electron microscopy (TEM), high-resolution transmission electron microscopy (HRTEM) and selected area electronic diffraction (SAED) and their pseudocapacitive properties were investigated in a 0.5 M aqueous Li2SO4 solution by cyclic voltammetry (CV), chronopotentiometry (CP) and AC impendence. The results show that the dimensions of the as-prepared α-MoO3 nanobelts are 200–400 nm in width, ca. 60 nm in thickness and 3–8 µm in length. The redox potential for the α-MoO3 nanobelts is found in the range of −0.3 to −1.0 V vs. SCE, which indicates that the α-MoO3 nanobelts can be used as anode electrode materials for hybrid supercapacitors. The specific capacitances of the α-MoO3 nanobelts at 0.1, 0.25, 0.5 and 1 A g−1 are 369, 326, 256 and 207 F g−1, respectively. The maximum specific capacitance of the α-MoO3 nanobelts is much higher than those of MoO3 nanoplates with 280 F g−1, MoO3 nanowires with 110 F g−1 and MoO3 nanorods with 30 F g−1 recently reported in literature. Furthermore, the α-MoO3 nanobelt electrode exhibits a good cycle stability with more than 95% of the initial specific capacitance maintained after 500 cycles. Additionally, the present route to prepare nanostructured MoO3 is much less expensive than those with Mo powders as the Mo source. Overall, the obtained high performance α-MoO3 nanobelts could be a promising electrode material for supercapacitors.
Motor imagery electroencephalogram (MI-EEG) classification is a vital task for brain computer interface (BCI) system. But the classification accuracy is not satisfactory, which hinders its generalizability. This study proposes a Filter-Bank Long Short-Term Memory (LSTM) Network (FBLSTM), which adopts a series of band-pass filters to obtain the different frequency information in EEG signals, and a convolution neural network to obtain spatial information. Moreover, a LSTM with attention mechanism is employed to process time series information. The open BCI Competition IV dataset 2a is applied to validate the performance of the proposed FBLSTM. Compared with recent methods, our method shows advantages on the within-subject and cross-subject 4-class classification performance and outperformed existing models, achieving an average accuracy of 72.4% and 53.6%, respectively.
Synthesis of cesium tungsten bronze (CsxWO3) nanorods with excellent near-infrared shielding performance by a one-step solvothermal method from cheap tungsten source is still challenging. In this work, small-sized CsxWO3 nanorods with excellent transparent thermal insulation performance were synthesized by a one-step solvothermal method. It was found that halogen acids played an important role in regulating the growth and microstructure of CsxWO3 nanorods. Moreover, different halogen acid can regulate the content of W5+=O and oxygen vacancies (VO) in CsxWO3 nanorods, which is helpful to further improve the near-infrared shielding performance of CsxWO3 particles. Compared with HF and HBr, the CsxWO3 nanoparticles synthesized under the regulation of HCl contain more VOs and W5+=O content, showing more excellent near-infrared shielding performance, and the as-prepared CsxWO3-PVA composite film exhibits low haze and high transparency, with the near-infrared light shielding rate at 1500 nm attaining to 97.7% when the visible light transmittance is 71.8%. The thermal insulation test results show that the CsxWO3-PVA composite film synthesized under the regulation of HCl can effectively reduce the indoor temperature by 6.9 °C under the irradiation of sunlight. After 100 cycles of thermal insulation tests, the near-infrared shielding and thermal insulation performance have not declined, showing excellent photothermal stability. This work is of great significance for promoting the industrial production of high-performance CsxWO3 nanoparticles and its application in energy-saving windows.
There is limited evidence linking exposure to ambient particulate matter (PM) with internal doses of metals and metalloids (metal(loid)s). This study aimed to evaluate the effects of short-term exposure to ambient PM on urine metal(loid)s among Chinese older adults. Biological monitoring data of 15 urine metal(loid)s collected in 3, 970 community-dwelling older adults in Fuyang city, Anhui Province, China, from July to September 2018, were utilized. PMs with an aerodynamic diameter ≤ 1 µm (PM1), ≤ 2.5 µm (PM2.5), and ≤ 10 µm (PM10) up to eight days before urine collection were estimated by space-time extremely randomized trees (STET) model. Residential greenness was reflected by Normalized Difference Vegetation Index (NDVI). We used generalized additive model (GAM) combined with distributed lag linear/non-linear models (DLMs/DLNMs) to estimate the associations between short-term PM exposure and urine metal(loid)s. The results suggested that the cumulative exposures to PM1, PM2.5, or PM10 over two days (lag0-1 days) before urine collection were associated with elevated urine metal(loid)s in DLMs, while exhibited linear or "inverted U-shaped" relationships with seven urine metal(loid)s in DLNMs, including Gallium (Ga), Arsenic (As), Aluminum (Al), Magnesium (Mg), Calcium (Ca), Uranium (U), and Barium (Ba). Aforementioned results indicated robust rather than spurious associations between PMs and these seven metal(loid)s. After standardizations for three PMs, PM1 was the greatest contributor to U, PM2.5 made the greatest contributions to Ga, As, Al, and Ba, and PM10 contributed the most to Mg and Ca. Furthermore, the effects of three PMs on urine Ga, As, Al, Mg, Ca, and Ba were reduced when exposed to higher levels of NDVI. Overall, short-term exposures to ambient PMs contribute to elevated urinary metal(loid) levels in older adults, and three PMs exhibit various contributions to different urine metal(loid)s. Moreover, residential greenness may attenuate the effects of PMs on urine metal(loid)s.
State-of-the-art spoken term detection or query-by-example networks depend on recurrent neural network (RNN), which extract fixed-dimensional vectors (embedded vectors) from both spoken query and the search content, and then calculate cosine distances over the vectors. However, these methods depend on time sequence, so it is a computational cost task, can not meet the requirements of both the query accuracy and search speed. In this work, we introduce a fast Spoken term detection system based on a separable model—RepVGG. Because of the trick of reparameterization, it has a faster speed in inference. Secondly, we use non maximum suppression and norm in the step of inference to improve it performance. Thirdly, we use multilanguage training to improve both accuracy and robustness of the system. Corresponding experiments are designed to verify these ideas. It show that proposed methods can import the GPU real-time factor (RTF) from 150 to 2300, and outperforms the state-of-the art method.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)