Enhanced wide-activated residual network for efficient and accurate image deblocking

To save bandwidth and storage space as well as speed up data transmission, people usually perform lossy compression on images. Although the JPEG standard is a simple and effective compression method, it usually introduces various visually unpleasing artifacts, especially the notorious blocking artifacts. In recent years, deep convolutional neural networks (CNNs) have seen remarkable development in compression artifacts reduction. Despite the excellent performance, most deep CNNs suffer from heavy computation due to very deep and wide architectures. In this paper, we propose an enhanced wide-activated residual network (EWARN) for efficient and accurate image deblocking. Specifically, we propose an enhanced wide-activated residual block (EWARB) as basic construction module. Our EWARB gives rise to larger activation width, better use of interdependencies among channels, and more informative and discriminative non-linearity activation features without more parameters than residual block (RB) and wide-activated residual block (WARB). Furthermore, we introduce an overlapping patches extraction and combination (OPEC) strategy into our network in a full convolution way, leading to large receptive field, enforced compatibility among adjacent blocks, and efficient deblocking. Extensive experiments demonstrate that our EWARN outperforms several state-of-the-art methods quantitatively and qualitatively with relatively small model size and less running time, achieving a good trade-off between performance and complexity.     

木质纤维与水泥共同改良软土的力学性能与微观机制分析

通过压实试验、无侧限抗压强度试验和劈裂试验,分析不同木质纤维含量、水泥含量和固化时间对软土力学性能的影响规律,探讨木质纤维、水泥改良软土的微观机制。结果表明,木质纤维的加入对水泥改良软土的击实特性有显著的影响;木质纤维与水泥可有效改善土体的抗压和劈裂抗拉强度,随着木质纤维含量的增加,改良土的抗压和劈裂抗拉强度呈现出明显的“驼峰”现象,并在木质纤维含量为0.25%时最大;木质纤维与水化产物、软土颗粒形成互锁效应,增大了改良土的摩擦力,同时木质纤维还承担一定的拉伸强度,使改良土的劈裂强度增加。     

Formation mechanism of porous reaction-bonded silicon nitride with interconnected pores in the presence of MgO

In porous reaction bonded silicon nitride, whiskers normally grow in globular clusters as the dominant morphology and deteriorate the pore interconnectivity. However, the ceramic microstructure was significantly transformed with the addition of MgO; specifically, the morphology was modified to a combination of matte and hexagonal grains. Microstructural observation along with thermodynamic studies suggest that MgO interfered with the presence and nitridation of SiO_(g). Consequently, rather than being involved in the whiskers’ formation, surface silica instead reacted with volatile MgO to form intermediate products. Through these reactions, whisker formation was blocked, and a porous interconnected structure formed which was confirmed by 3D tomography. After heat-treatment at 1700 °C, β-Si₃N₄ crystallized in a glassy matrix containing magnesium. Resulting samples had an open-pore structure with porosity of 74–84 vol. %, and density of 0.48-0.75 g.cm^?3. Combination of high porosity and pore size of <40 μm led to compressive strengths of 1.1–1.6 MPa.     

海藻多酚功能性作用机制及其应用初探

海藻多酚是海藻次级代谢产物,具有多种生物学活性,能够为人体所吸收和利用,所以海藻多酚成为了现阶段海洋活性物质研究的一个重点。就海藻多酚功能性作用机制以及其应用做分析,旨在为实践提供指导和帮助。     

Carbon corrosion mechanism on nitrogen-doped carbon support — A density functional theory study

In this work, density functional theory (DFT) calculations were used to investigate the mechanism of carbon corrosion on nitrogen-doped carbon support. Free energy diagrams were generated based on three proposed reaction pathways to evaluate corrosion mechanisms. The most energetically preferred mechanism on nitrogen-doped carbon was determined. The results show that the step of water dissociation to form ^#OH was the rate-determining step for gra-G-1N (graphene doped with graphitic N) and pyrr-G-1N (graphene doped with pyrrolic N). As for graphene doped with pyridinic N, the step of C^#OC^#O formation was critical. It was found that the control of nitrogen concentration was necessary for precisely designing optimized carbon materials. Abundance of nitrogen moieties aggravated the carbon corrosion. When the high potential was applied, specific types of graphitic N and pyridinic N were found to be favorable carbon modifications to improve carbon corrosion resistance. Moreover, the solvent effect was also investigated. The results provide theoretical insights and design guidelines to improve corrosion resistance in carbon support through material modification by inhibiting the adsorption of surface oxides (OH, O, and OOH).     

Simple and novel synthesis of zirconia whiskers from a phosphate flux

High quality zirconia whiskers have been successfully prepared by molten salt method, using zirconium oxychloride (ZrOCl₂·8H₂O) and sodium phosphate tribasic dodecahydrate (Na₃PO₄·12H₂O) as precursor and molten salt, respectively. The effects of types of molten salt and heat treatment temperature on the formation of zirconia whiskers were characterized by XRD, Raman, DTA-TG, FE-SEM, TEM, SAED and HR-TEM. When Na₃PO₄·12H₂O is utilized as molten salt and the heat treatment temperature is 900?°C, the as-prepared zirconia whiskers with length ranging from 4?µm to 8?µm show an average aspect ratio of 25. The obtained ZrO₂ whiskers with monoclinic structure are elongated along [010] direction and exhibit a smooth surface with no distinct defects. The XRD and Raman results reveal that the phase transformation from tetragonal zirconia to monoclinic zirconia occurs with the increased crystal size and the water quenching treatment can significantly reduce the content of sodium zirconium phosphate [Na_9–4×Zr_x(PO₄)₃] in the final product. The growth mechanism of zirconia whiskers is supposed to be a dissolution-precipitation process. Since the sodium zirconium phosphate [Na_9–4×Zr_x(PO₄)₃] effectively promotes the dissolution of zirconia in liquid molten salt, zirconia can grow into zirconia whiskers according to its anisotropy.     

2-Hydroxy-4-methoxybenzaldehyde inhibits the growth of Aspergillus flavus via damaging cell wall,cell membrane,manipulating respiration thus creating a promising antifungal effect on corn kernels

This study investigated the antifungal activity and the potential antifungal mechanisms of 2-hydroxy-4-methoxybenzaldehyde (HMB) against Aspergillus flavus (A. flavus). The minimum inhibitory concentration (MIC) of HMB in preventing spore germination was 70 μg mL⁻¹. HMB at MIC disrupted cell wall integrity by reducing the number of septa by 86.66% (P < 0.05) in mycelia and increased cell membrane permeability by about 14-fold (P < 0.05) evidenced by propidium iodide (PI) staining. Furthermore, HMB at MIC inhibited respiration by 33.33%. These results revealed that the antifungal activity of HMB against A. flavus could be attributed to the damaged cell wall integrity, cell membrane permeability and respiration metabolism. What’s more, A. flavus was completely restrained in corn kernels due to HMB. Therefore, HMB could be applied as an effective antifungal agent.     

Synthesis and photocatalytic activity of g-C3N4/ZnO composite microspheres under visible light exposure

In this paper, a novel g-C₃N₄/ZnO composite microspheres (CZCM) with enhanced photocatalytic activity under visible light exposure were successfully prepared by a self-assembly method followed by calcination in the air. A hierarchical structure in which ZnO microspheres were closely covered with g-C₃N₄ nanosheets was constructed. The microstructure and photocatalytic activities of the CZCM were characterized. The photocatalytic property of CZCM was evaluated by degrading solution Methyl Orange (MO) and Tetracycline (TC). The effects of varied contents of g-C₃N₄ on the photocatalytic capability of CZCM were systematically investigated and the results show that the optimized CZ-15% sample exhibit much higher photocatalytic degradation efficiency than that of bare g-C₃N₄ or ZnO under identical conditions. The analysis of Photoluminescence (PL) and photocurrent (PC) independently conformed that the photo-induced electron-hole (e^?-h⁺) pairs in the CZCM were effectively generated and responsible for the observed photocatalysis. The enhanced adsorption of visible-light and the effective charge separation on the surface of CZCM enabled significant improvement of photocatalytic performance. According to the experimental results and relative energy band levels of the two semiconductors, a possible photocatalysis mechanism for the reaction process is proposed.     

High-entropy ferroelastic rare-earth tantalite ceramic: (Y0.2Ce0.2Sm0.2Gd0.2Dy0.2)TaO4

In this study, high-entropy rare-earth tantalate ceramics (Y_0.2Ce_0.2Sm_0.2Gd_0.2Dy_0.2)TaO₄ ((5RE_0.2)TaO₄) have been successfully fabricated. The possibility of formation of (5RE_0.2)TaO₄ was verified via first-principles calculations. In addition, the phase structure, ferroelastic toughening mechanism, thermophysical, and mechanical properties were systematically investigated. The (5RE_0.2)TaO₄ ceramics have lower phonon thermal conductivity (1.2–2.6 W·m^–1·K^–1) in the entire temperature range than that of RETaO₄ and YSZ. (5RE_0.2)TaO₄ has a higher fracture toughness and lower brittleness index than YSZ. The thermal expansion coefficients of (5RE_0.2)TaO₄ are as high as 10.3 × 10^-6 K^–1 at 1200°C and Young's modulus is 66–189 GPa, and thus, (5RE_0.2)TaO₄ possesses great potential for application in thermal barrier coatings (TBCs).     

Sequential alignment attention model for scene text recognition

Scene text recognition has been a hot research topic in computer vision due to its various applications. The state-of-the-art solutions usually depend on the attention-based encoder-decoder framework that learns the mapping between input images and output sequences in a purely data-driven way. Unfortunately, there often exists severe misalignment between feature areas and text labels in real-world scenarios. To address this problem, this paper proposes a sequential alignment attention model to enhance the alignment between input images and output character sequences. In this model, an attention gated recurrent unit (AGRU) is first devised to distinguish the text and background regions, and further extract the localized features focusing on sequential text regions. Furthermore, CTC guided decoding strategy is integrated into the popular attention-based decoder, which not only helps to boost the convergence of the training but also enhances the well-aligned sequence recognition. Extensive experiments on various benchmarks, including the IIIT5k, SVT, and ICDAR datasets, show that our method substantially outperforms the state-of-the-art methods.