Application of Self-Attention Generative Adversarial Network for Electromagnetic Imaging in Half-Space.

In this paper, we introduce a novel artificial intelligence technique with an attention mechanism for half-space electromagnetic imaging. A dielectric object in half-space is illuminated by TM (transverse magnetic) waves. Since measurements can only be made in the upper space, the measurement angle will be limited. As a result, we apply a back-propagation scheme (BPS) to generate an initial guessed image from the measured scattered fields for scatterer buried in the lower half-space. This process can effectively reduce the high nonlinearity of the inverse scattering problem. We further input the guessed images into the generative adversarial network (GAN) and the self-attention generative adversarial network (SAGAN), respectively, to compare the reconstruction performance. Numerical results prove that both SAGAN and GAN can reconstruct dielectric objects and the MNIST dataset under same measurement conditions. Our analysis also reveals that SAGAN is able to reconstruct electromagnetic images more accurately and efficiently than GAN.

Dual-cell structure digitally controlled oscillator with portability for clock generation applications.

This paper presents a dual-cell structure and cell-based design digitally controlled oscillator (DCO) for high-performance clock generation applications. The proposed dual-cell DCO not only can reduce the intrinsic delay to increase the maximum output frequency significantly but also extend the controllable range. In addition, the proposed digitally controlled delay cell uses the cascading structure to realize the wide output frequency range and high timing resolution at the same time. A DCO chip is fabricated in 0.18 µm CMOS technology, and the core area is 0.003 mm2. Measurement results show that operation range is from 169 MHz to 522 MHz and the finest delay resolution is 2.2 ps. Furthermore, because the proposed DCO is implemented with the digital standard cells, it is a portable design to migrate to different processes easily and reduce design time significantly.

Effect of lactic acid bacteria isolated from fermented mustard on immunopotentiating activity

OBJECTIVE: To investigate the effect of lactic acid bacteria isolated from fermented mustard on immunopotentiating activity METHODS: One hundred and fifty nine strains of lactic acid bacteria isolated from traditional Taiwan fermented mustard were evaluated for their immunopotentiating activity on a murine macrophage cell line RAW 264.7. RESULTS: Of the strains, pronounced increases in the levels of nitric oxide (NO), tumor necrosis factor-α and interleukin-6 were observed in strains B0040, B0110 and B0145. Among them, strain B0145 had the highest NO and tumor necrosis factor-α generation in RAW 264.7 cells; strains B0040 and B0110 were also superior to that of Lactobacillus casei. These results demonstrated that NO and cytokines were effectively induced when the bacterial stimulants were treated with macrophages. In addition, strains B0040 and B0110 were identified as Lactobacillus plantarum, and B0145 as Weissella cibaria using 16S rDNA analysis. CONCLUSIONS: The results implicated selected strains may be regarded as a biological response modifier and had a broad application prospects in exploiting new functional food or as a feed additive.