Text-to-image synthesis aims to generate realistic images conditioned on text descriptions. To fuse text information into synthesized images, conditional affine transformations (CATs), such as conditional batch normalization (CBN) and conditional instance normalization (CIN), are usually used to predict batch statistics of different layers. However, ordinary CAT blocks control the batch statistics independently disregarding the consistency among neighboring layers. To address the above issue, we propose a new fusion approach names recurrent affine transformation (RAT) for synthesizing images conditioned on text information. RAT connects all the CAT blocks with recurrent connections for explicitly fitting the temporal consistency between CAT blocks. To verify the effectiveness of RAT, we propose a novel visualization method to show how generative adversarial network (GAN) fuses conditional information. Our microscopic and macroscopic visualizations not only demonstrate the effectiveness of RAT but also turn out to be a useful perspective to analyze how GAN fuses conditional information. In addition, we propose a more stable spatial attention mechanism for the discriminator, which helps the text description to supervise the generator to synthesize more relevant image contents. Extensive experiments on the CUB, Oxford-102, and COCO datasets demonstrate the proposed model's superiority in comparison to state-of-the-art models. Our code is available on <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/senmaoy/RAT-GAN</uri> .

All-natural rubber is harvested from rubber trees (Hevea brasiliensis Muell. Arg.) by traditional tapping knives, so rubber tapping still heavily relies on labor. Therefore, this study explored a novel, hand-held mechanical rubber tapping machine for rubber tree harvesting. In this study, a mechanical tapping cutter with a vertical blade and adjustable guide was first described. The response surface method was applied to evaluate factors affecting the tapping effect. The experimental values were in close agreement with the predicted value. Machine-tapped latex was comparable in quality to hand-tapped latex. Based on the single-factor results, the response surface method (RSM) and the center combined rotation design (CCRD) optimization method were adopted to explore the influence of three factors influencing vertical blade height (A), cutting force (B), and spiral angle (C) on the tapping effect. Regarding the cutting rate of the old rubber line (Y1), cutting time (Y2), latex flow rate (Y3), and average cutting current (Y4) as evaluation indexes of the tapping effect, an optimization scheme was determined. The quadratic model fits for all the responses. The test results showed that the main factors affecting Y1, Y2, Y3, and Y4 were A and B, B, A and C, and B, respectively. Under optimal conditions, the influencing factors of A, B, and C were 10.24 mm, 51.67 N, and 24.77°, respectively, when the evaluation index values of Y1, Y2, Y3, and Y4 were 98%, 8.65 mL/5 min, 9.00 s, and 1.16 A. The range of the relative error between the experimental and predicted results was from −11.11% to 11.11%. According to the optimized treatment scheme, a comparison test was designed between mechanical and manual rubber tapping tools. To verify the availability and effect of the mechanical tapping method preliminarily, the important rubber tapping evaluation indexes included bark thickness, bark excision, latex flow time, cutting time, ash content, and cutting depth, which were selected to serve as a comparison test. There was no significant difference between hand and mechanical methods, except ash content (p &lt; 0.05) and cutting time (p &lt; 0.01). The mechanical tapping machine proposed in this study is meaningful to improve cutting efficiency, practicality, and operability. Furthermore, it provides crucial theoretical references for the development of intelligent tapping machines.

Abstract Studies have shown that exposure to either fungicide difenoconazole (DIF) or salt (NaCl) alone can cause phytotoxicity in plants, but it remains uncertain whether co-exposure to these two xenobiotics, which often co-occur in the agroecosystem, can also induce phytotoxicity. This research aimed to study the phytotoxicity and underlying mechanisms of co-exposure to DIF (field recommended doses (0.5 L ha− 1) and NaCl (150 mM) in tomato seedlings. The results showed that exposure to DIF and/or NaCl causes phytotoxic effects in tomato seedlings, including a decrease in fresh weight and length of shoots and roots, damage to chlorophyll pigment, and induced oxidative stress in the leaves. Interestingly, combined exposure to DIF and NaCl exhibited synergistic effects on shoot and root biomass inhibition. Antioxidant defense analyses revealed that the DIF and/or NaCl exposure altered the activities of enzymes involved in the H2O2 scavenging (ascorbate peroxidase and catalase), and in xenobiotic detoxification (glutathione-s-transferase and peroxidase) in leaf tissues. Interestingly, combined exposure to DIF and NaCl markedly enhanced detoxifying enzymes, thereby enhancing xenobiotic biotransformation. In addition, DIF and/or NaCl exposure enhanced proline accumulation, and altered the thiols profile (reduced glutathione) content, while stimulating the phenylpropanoid pathway (phenylalanine ammonia-lyase activity) to produce secondary metabolites such as polyphenols and flavonoids. As pioneer research to highlight the phytotoxicity induced by co-exposure to DIF and NaCl in tomato seedlings, these results provide new insights into physio-biochemical responses of non-target plants to DIF and NaCl co-contamination and shedding light on the ecological risks of pesticides and salt exposure in agroecosystems.

Gelsegansymines A (1) and B (2), two new indole alkaloids along with six known analogues (3-8) were isolated from the aerial parts of Gelsemium elegans. Their structures were elucidated by means of spectroscopic techniques. Structurally, compounds 1 and 2 possessed the rare cage-like gelsedine skeleton hybrid with bicyclic monoterpenoid. The anti-inflammatory activities of isolated compounds (1-3) were tested on LPS induced RAW264.7 cells. Under the treated concentration without toxicity for cells, the cytokines levels of nitric oxide (NO), tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6) were evaluated by Griess method and enzyme-linked immunosorbent assay (ELISA). The results showed that compounds 1-3 exhibited anti-inflammatory activities with dose-dependent manner range from 12.5 to 50 μmol/L. Furthermore, the inhibitory activities of compounds 1 and 2 on receptor activator of NF-κB ligand (RANKL) induced osteoclast formation were tested in vitro. Compounds 1 and 2 at 5 μmol/L exhibited the significant inhibitory effect on the osteoclastogenesis induced by RANKL. This work reported the anti-inflammatory and osteoclast inhibitory activities of new monoterpenoid indole hybrids, which may inspire the further light on the related traditional application research of G. elegans.

In order to find better orthopedic implant materials and structures, three triply periodical minimal surface (TPMS) structures using Ti64-5Cu alloy of Gyroid (G), Diamond (D) and Primitive (P) with fixed porosity were prepared by laser powder bed fusion (LPBF). The permeability and flow properties of the three TPMS structures were analyzed based on hydrodynamics simulations, the corrosion resistance and antibacterial properties were compared with those using Ti64 alloy. The results shows that the corrosion current (15.85 μA/cm2) of Ti64-5Cu is significantly lower than that (45.70 μA/cm2) of Ti64, and the antibacterial performance is significantly improved. The material transportation ability of the G structure is 42% higher than D structure. So the G-type TPMS structure based on Ti64-5Cu material has performance as an orthopedic implant in terms of meeting the growth of human bone tissue cells and bone modulus adaptation.

Preferential flow benefits water storage in deep soils that contribute to hydrological sustainability under different vegetations in the Karst regions, however, how vegetation specializes preferential flow pattern by changing soil architecture remains unclear. The study investigated the variations of soil architecture and preferential flow patterns in the agricultural field, abandoned field, and young and mature woodlands that representing four vegetation succession stages during the implementation of returning cultivated land into forest in the Karst region, to explored how soil texture and structure interact to specialize preferential flow patterns under the vegetation succession, using the dye-tracing technique, the principal component analysis (PCA), and the variation partitioning analysis (VPA). The dyed soil profiles illustrated heavy preferential flow under the four vegetations. The maximum infiltration depth (Dmax), actual infiltration depth (Dactual), total dye coverage (Ctotal), and length index (LI) significantly increased following the sequence of vegetation succession stages (P < 0.05), and the preferential pathway contribution (Dpr) was the significantly lowest under the agricultural field and young woodland and the highest under the mature woodland (P < 0.05), which indicated of intensified preferential flow with ongoing vegetation succession. The preferential flow pattern index (PFP) was constructed by the Dmax, Dactual, Ctotal, Dpr, and LI after the PCA, which gradually increased (from −0.82 ± 0.69 to 1.37±0.80) during the succession, suggesting that preferential flow percolated deeper, more preferential flow was received, and preferential flow contribution increased during the succession. Soil texture (sand content, silt content, and geometric mean soil particle size) and structure (bulk density, air-filled porosity, total porosity, air-filled porosity, and ratio of air-filled and capillary porosity to total porosity) significantly correlated to the PFP (P < 0.05). The VPA revealed that soil texture and structure independently explained 9 % and 26 % variation of the PFP. This situation means although reforestation intensified preferential flow by improving soil structure, adverse impacts of improper soil texture management of could offset the intensification efficiency by reforestation. Therefore, the results benefit to meet ecological and hydrological sustainability challenges rooted by preferential flow in the Karst regions.

The article presents the results of the analysis of statistical data on the functioning of the system of science cities, as well as the results of monitoring the performance of science cities, the study of financial tools for supporting science cities. Based on the results of the analysis, proposals were formulated to improve the measures of state support for science cities.

With the increasing demand for wheat, the detection of wheat quality has become imperative. Protein content is an important indicator for wheat quality. Near infrared spectroscopy (NIRS) quantitative non-destructive testing technology has gained widespread application in agricultural field with the development of science and chemometrics technology. In this study, NIRS system was employed to measure the spectra of wheat, and the original spectra were pretreated using Savitzky-Golay smoothing (SG) pretreatment method. Subsequently, the NIRS prediction model of protein in wheat that using SG combined with parallel convolutional neural network (PaBATunNet) was established. PaBATunNet was composed of a one-dimensional convolutional layer, a parallel convolution module (Module), a flattening layer, four fully connected layers and a parameter regulator (PR). Module was made up of five submodules and a Concatenate function. The multidimensional features of the spectra were extracted by five submodules and spliced by Concatenate function. SG pretreatment combined with PaBATunNet (SG-PaBATunNet) was compared with commonly modeling methods, such as SG-partial least squares (SG-PLS), SG-principal component regression (SG-PCR), SG-support vector machine (SG-SVM) and SG-back propagation neural network (SG-BP). The results demonstrated that the modeling accuracy and prediction accuracy of SG-PaBATunNet were improved by 26.7%, 23.9%, 45.6%, 44.2%, and 38.4%, 39.6%, 60.1%, 58.0%, when compared with SG-PLS, SG-PCR, SG-SVM and SG-BP. The problems of low prediction accuracy and poor generalization ability with commonly modeling methods were effectively addressed by SG-PaBATunNet. This study provides an essential theoretical foundation for developing a fast, nondestructive and high-precision NIRS quantitative analysis model of protein in wheat.