Aiming at the problem of radar emitter individual recognition with unintentional modulation feature, a method of emitter fingerprint identification based on deep learning is proposed. In this method, the deep fingerpr...
详细信息
The water states anddistribution in frozen dough are critical to its quality. In this study, variable temperature 1H time-domain nuclear magnetic resonance (VT-NMr) was used to in situ investigate the unfrozen water ...
详细信息
In this paper, different waterproof materials are used to brush the concrete surface, and the influence of different coating materials on the microstructure, porosity, mechanics anddurability of concrete is analysed....
详细信息
The thermal expansion property of epoxy mortar has a significant effect on the repair effect. Experiments were carried out to study the thermal expansion coefficient of epoxy mortar by changing the type of curing agen...
详细信息
A geomimetic, ultrafast, and one-step preparation method for an N-dopedreduced TiO2porous layer using the atmospheric plasma spraying (APS) technique is proposed. The secondary gas provided by the APS (H2) is respons...
详细信息
A geomimetic, ultrafast, and one-step preparation method for an N-dopedreduced TiO2porous layer using the atmospheric plasma spraying (APS) technique is proposed. The secondary gas provided by the APS (H2) is responsible forreducing TiO2, while an acetamide solution stored in the pores of the substrate acted as a nitrogen source for TiO2doping. Inspired by the nature of volcanic eruptions on Earth, we used a geomimetic vapor-induced pore-forming (VI-PF) mechanism to produce porous N-doped TiO2via one-step preparation. We varied the acetamide solution concentration to synthesize different samples (TNX,Xindicates the wt % of acetamide solution) to approach photocatalytic tetracycline (TC, 20 ppm/100 mL) degradation under AM1.5G solar simulator irradiation. When the acetamide solution concentration was 5 wt %, the sample (TN5) showed the highest lattice-nitrogen content, which gave it the highest photocatalytic degradation efficiency (rate constant = 1.71 × 10–2min–1) of TC. Moreover, after 15 rounds of recycling tests, the removal rate and crystalline structure remained stable. In addition to experiments, density functional theory (dFT) calculations are also used to verify the experimental results, and they are also consistent with each other. This work is the first study using the APS technique in a single-step process for nitrogen doping and H2reduction and is a green, cost-effective, time-saving, and promising method for the mass production of porous TiO2.
A series of novel EGFr inhibitors were designed and synthesized based on the scaffold of osimertinib, the inhibitory activities against the L858r/T790M/C797S mutant EGFr kinase of which were subsequently evaluated. Co...
详细信息
Geosynthetics have a great development prospect in seepage prevention of waterretaining buildings because of theirreliability and economic superiority. Geomembrane impermeable rockfill dam development at home and ab...
详细信息
regular through-pore membranes exhibit extremely high productivity and size-selective capabilities. Previous studies have demonstrated that the use of a polymer alone on a solid substrate surface hinders the formation...
详细信息
Many models combining Transformers with convolutional neural networks (CNNs) for computer vision tasks have achieved state-of-the-art results. However, due to the different computation patterns between attention and c...
Many models combining Transformers with convolutional neural networks (CNNs) for computer vision tasks have achieved state-of-the-art results. However, due to the different computation patterns between attention and convolution, using a dedicated Transformer or CNN accelerator will inevitably reduce the computing efficiency of the other. To overcome this problem, we propose a unified architecture for attention and convolution on FPGA. We reduce runtime overhead by offloading part of self-attention computations offline before inference. Furthermore, we present a unified mapping method according to the computing characteristics of attention-based and convolution-based models. This accelerator implements multi-head attention in Transformer, independent resNet-50 and hybrid blocks of attention and con-volution in BoTNet-50 at 200MHz on Xilinx Virtex Ultrascale+ XCVU37P. Experimental results show that the solution is nearly 3.62 times more energy-efficient than the NVIdIA V100 GPU, and the computational efficiency is 11.86% and 28.29% higher than the state-of-the-art Transformer andresNet-50 accelerators, respectively.
Here, we demonstrate world’s first 1 Hz driving fringe-field switching (FFS) LCd via development of low flexoelectric coefficient LCs with high voltage holding ratio (VHr). The evaluation of the 1 Hz driving 13.4-inc...
详细信息
暂无评论