In small-plot experiments, weed scientists have traditionally estimated herbicide efficacy through visual assessments or manual counts with wooden frames—methods that are time-consuming, labor-intensive, and error-prone. This study introduces a novel mobile application (app) powered by convolutional neural networks (CNNs) to automate the evaluation of weed coverage in turfgrass. The mobile app automatically segments input images into 10 by 10 grid cells. A comparative analysis of EfficientNet, MobileNetV3, MobileOne, ResNet, ResNeXt, ShuffleNetV1, and ShuffleNetV2 was conducted to identify weed-infested grid cells and calculate weed coverage in bahiagrass (Paspalum notatum Flueggé), dormant bermudagrass [Cynodon dactylon (L.) Pers.], and perennial ryegrass (Lolium perenne L.). Results showed that EfficientNet and MobileOne outperformed other models in detecting weeds growing in bahiagrass, achieving an F1 score of 0.988. For dormant bermudagrass, ResNet performed best, with an F1 score of 0.996. Additionally, app-based coverage estimates (11%) were highly consistent with manual assessments (11%), showing no significant difference (P = 0.3560). Similarly, ResNeXt achieved the highest F1 score of 0.996 for detecting weeds growing in perennial ryegrass, with app-based and manual coverage estimates also closely aligned at 10% (P = 0.1340). High F1 scores across all turfgrass types demonstrate the models’ ability to accurately replicate manual assessments, which is essential for herbicide efficacy trials requiring precise weed coverage data. Moreover, the time for weed assessment was compared, revealing that manual counting with 10 by 10 wooden frames took an average of 39.25, 37.25, and 42.25 s per instance for bahiagrass, dormant bermudagrass, and perennial ryegrass, respectively, whereas the app-based approach reduced the assessment times to 8.23, 7.75, and 14.96 s, respectively. These results highlight the potential of deep learning–based mobile tools for fast, accurate, scalable weed coverage assessments, enabling efficient herbicide trials and offering labor and cost savings for researchers and turfgrass managers.

Automatic precision herbicide application offers significant potential for reducing herbicide use in turfgrass weed management. However, developing accurate and reliable neural network models is crucial for achieving optimal precision weed control. The reported neural network models in previous research have been limited by specific geographic regions, weed species, and turfgrass management practices, restricting their broader applicability. The objective of this research was to evaluate the feasibility of deploying a single, robust model for weed classification across a diverse range of weed species, considering variations in species, ecotypes, densities, and growth stages in bermudagrass turfgrass systems across different regions in both China and the United States. Among the models tested, ResNeXt152 emerged as the top performer, demonstrating strong weed detection capabilities across 24 geographic locations and effectively identifying 14 weed species under varied conditions. Notably, the ResNeXt152 model achieved an F1 score and recall exceeding 0.99 across multiple testing scenarios, with a Matthews correlation coefficient (MCC) value surpassing 0.98, indicating its high effectiveness and reliability. These findings suggest that a single neural network model can reliably detect a wide range of weed species in diverse turf regimes, significantly reducing the costs associated with model training and confirming the feasibility of using one model for precision weed control across different turf settings and broad geographic regions.

With a security alliance with the United States and deep economic relations with China, South Korea faces complex foreign policy choices amid US–China competition. A critical decision is whether to join the Quadrilateral Security Dialogue (Quad), a US-led grouping widely viewed as aiming to counter China in the Indo-Pacific. The choice depends on its domestic politics as much as its relationships with both superpowers. Using a public opinion survey with a priming experiment, we investigate South Korean citizens’ preferences regarding the Quad. We find that, without additional information, nearly half of the respondents supported joining the Quad. Yet neither mentioning the security benefits of joining the Quad nor mentioning the potential economic costs associated with Chinese retaliation for joining the Quad changed their level of support. Nor did we detect any treatment heterogeneity. Beyond the experiment, we find that threat perceptions and party affiliation are strongly correlated with respondents’ preferences.

Folate metabolism is involved in the development and progression of various cancers. We investigated the association of single nucleotide polymorphisms (SNP) in folate-metabolising genes and their interactions with serum folate concentrations with overall survival (OS) and liver cancer-specific survival (LCSS) of newly diagnosed hepatocellular carcinoma (HCC) patients. We detected the genotypes of six SNP in three genes related to folate metabolism: methylenetetrahydrofolate reductase (MTHFR), 5-methyltetrahydrofolate-homocysteine methyltransferase reductase (MTRR) and 5-methyltetrahydrofolate-homocysteine methyltransferase (MTR). Cox proportional hazard models were used to calculate multivariable-adjusted hazard ratios (HR) and 95 % CI. This analysis included 970 HCC patients with genotypes of six SNP, and 864 of them had serum folate measurements. During a median follow-up of 722 d, 393 deaths occurred, with 360 attributed to HCC. In the fully-adjusted models, the MTRR rs1801394 polymorphism was significantly associated with OS in additive (per G allele: HR = 0·84, 95 % CI: 0·71, 0·99), co-dominant (AG v. AA: HR = 0·77; 95 % CI: 0·62, 0·96) and dominant (AG + GG v. AA: HR = 0·78; 95 % CI: 0·63, 0·96) models. Carrying increasing numbers of protective alleles was linked to better LCSS (HR10–12 v. 2–6 = 0·70; 95 % CI: 0·49, 1·00) and OS (HR10–12 v. 2–6 = 0·67; 95 % CI: 0·47, 0·95). Furthermore, we observed significant interactions on both multiplicative and additive scales between serum folate levels and MTRR rs1801394 polymorphism. Carrying the variant G allele of the MTRR rs1801394 is associated with better HCC prognosis and may enhance the favourable association between higher serum folate levels and improved survival among HCC patients.

The elasto-inertial focusing and rotating characteristics of spheroids in a square channel flow of Oldroyd-B viscoelastic fluids are studied by the direct forcing/fictitious domain method. The rotational behaviours, changes in the equilibrium positions and travel distances are explored to analyse the mechanisms of spheroid migration in viscoelastic fluids. Within the present simulated parameters (1 ≤ Re ≤ 100, 0 ≤ Wi ≤ 2, 0.4 ≤ α ≤3), the results show that there are four kinds of equilibrium positions and six (five) kinds of rotational behaviours for the elasto-inertial migration of prolate (oblate) spheroids. We are the first to identify a new rotational mode for the migration of prolate spheroids. Only when the particles are initially located at a corner and wall bisector, some special initial orientations of the spheroids have an impact on the final equilibrium position and rotational mode. In other general initial positions, the initial orientation of the spheroid has a negligible effect. A higher Weissenberg number means the faster the particles migrate to the equilibrium position. The spheroid gradually changes from the corner (CO), channel centreline (CC), diagonal line (DL) and cross-section midline (CSM) equilibrium positions as the elastic number decreases, depending on the aspect ratio, initial orientation and rotational behaviour of the particles and the elastic number of the fluid. When the elastic number is less than the critical value, the types of rotational modes of the spheroids are reduced. By controlling the elastic number near the critical value, spheroids with different aspect ratios can be efficiently separated.

As a required sample preparation method for 14C graphite, the Zn-Fe reduction method has been widely used in various laboratories. However, there is still insufficient research to improve the efficiency of graphite synthesis, reduce modern carbon contamination, and test other condition methodologies at Guangxi Normal University (GXNU). In this work, the experimental parameters, such as the reduction temperature, reaction time, reagent dose, Fe powder pretreatment, and other factors, in the Zn-Fe flame sealing reduction method for 14C graphite samples were explored and determined. The background induced by the sample preparation process was (2.06 ± 0.55) × 10–15, while the 12C– beam current were better than 40μA. The results provide essential instructions for preparing 14C graphite of ∼1 mg at the GXNU lab and technical support for the development of 14C dating and tracing, contributing to biology and environmental science.

A new vacuum line to extract CO2 from carbonate and dissolved inorganic carbon (DIC) in water was established at Guangxi Normal University. The vacuum line consisted of two main components: a CO2 bubble circulation region and a CO2 purification collection region, both of which were made of quartz glass and metal pipelines. To validate its reliability, a series of carbonate samples were prepared using this system. The total recovery rate of CO2 extraction and graphitization exceeded 80%. Furthermore, the carbon content in calcium carbonate exhibited a linear relationship with the CO2 pressure within the system, demonstrating its stability and reliability. The system was also employed to prepare and analyze various samples, including calcium carbonate blanks, foraminiferal, shell, groundwater, and subsurface oil-water samples. The accelerator mass spectrometry (AMS) results indicated that the average beam current for 12C- in the samples exceeded 40 μA. Additionally, the contamination introduced during the liquid sample preparation process was approximately (1.77 ± 0.57) × 10−14. Overall, the graphitized preparation system for carbonate and DIC in water exhibited high efficiency and recovery, meeting the requirements for samples dating back to approximately 30,000 years.

This article investigates how superpower rivalry affects public perceptions of international organization (IO) legitimacy in the hegemon. We argue that the representation of a superpower rival state at an IO in the form of its key decision maker's nationality can dampen the IO's perceived legitimacy within the rival power. We test this argument using a survey experiment in the United States under President Trump, where we manipulate the nationality of the International Court of Justice (ICJ) judge who casts a tie-breaking vote against the United States. Our results show that when the judge is Chinese, there is a strong and robust dampening of Americans’ perceptions of the ICJ's legitimacy, with no comparable effect arising when the judge is from other countries, including Russia. Replication of the experiment in the United States under President Biden offers external validity for our findings, which may have important implications for the future of the liberal international order.

Censorship is one of the main forms of political coercion deployed by modern states to control and regulate public expression. In this article, we examine the political censorship of China’s intellectual public space, which has long been underexplored. We apply unsupervised machine learning to examine the database of a leading intellectual portal website, which serves as an archive of both published and censored intellectual writings between 2000 and 2020 and includes over 740 million Chinese characters. We identify a strategic censorship mechanism that consists of thematic and persona censorship elements. Thematic censorship involves the state filtering out writing that competes with the official policy narrative, historiography, and values. Persona censorship involves the complete muting of individual intellectuals who have previously made derogatory attacks on the supreme leaders of the Communist Party, which represents a symbolic act of open defiance.

Sterol regulatory element-binding protein 2 (SREBP2) is considered to be a major regulator to control cholesterol homoeostasis in mammals. However, the role of SREBP2 in teleost remains poorly understand. Here, we explored the molecular characterisation of SREBP2 and identified SREBP2 as a key modulator for 3-hydroxy-3-methylglutaryl-coenzyme A reductase and 7-dehydrocholesterol reductase, which were rate-limiting enzymes of cholesterol biosynthesis. Moreover, dietary palm oil in vivo or palmitic acid (PA) treatment in vitro elevated cholesterol content through triggering SREBP2-mediated cholesterol biosynthesis in large yellow croaker. Furthermore, our results also found that PA-induced activation of SREBP2 was dependent on the stimulating of endoplasmic reticulum stress (ERS) in croaker myocytes and inhibition of ERS by 4-Phenylbutyric acid alleviated PA-induced SREBP2 activation and cholesterol biosynthesis. In summary, our findings reveal a novel insight for understanding the role of SREBP2 in the regulation of cholesterol metabolism in fish and may deepen the link between dietary fatty acid and cholesterol biosynthesis.

Compared with nitrogen and argon, helium is lighter and can better reduce the beam loss caused by angular scattering during beam transmission. The molecular dissociation cross-section in helium is high and stable at low energies, which makes helium the prevalent stripping gas in low-energy accelerator mass spectrometry (AMS). To study the stripping behavior of 14C ions in helium at low energies, the charge state distributions of carbon ion beams with −1, +1, +2, +3, and +4 charge states were measured at energies of 70–220 keV with a compact 14C-AMS at Guangxi Normal University (GXNU). The experimental data were used to analyze the stripping characteristics of C-He in the energy range of 70–220 keV, and new charge state yields and exchange cross-sections in C-He were obtained at energies of 70–220 keV.

A single-stage accelerator mass spectrometer (GXNU-AMS) developed for radiocarbon and tritium measurements was installed and commissioned at Guangxi Normal University in 2017. After several years of operational and methodological upgrades, its performance has been continuously improved and applied in multidisciplinary fields. Currently, the measurement sensitivity for radiocarbon and tritium is 14C/12C ∼ (3.14 ± 0.05) ×10–15 and 3H/1H ∼ (1.23 ± 0.17)×10–16, respectively, and the measurement accuracy is ∼0.6%, which can meet the measurement requirements in the nuclear, earth, environmental and life science fields. This study presents the performance characteristics of GXNU-AMS and several interesting application studies.

A single-shot measurement of electron emittance was experimentally accomplished using a focused transfer line with a dipole. The betatron phase of electrons based on laser wakefield acceleration (LWFA) is energy dependent owing to the coupling of the longitudinal acceleration field and the transverse focusing (defocusing) field in the bubble. The phase space presents slice information after phase compensation relative to the center energy. Fitting the transverse size of the electron beam at different energy slices in the energy spectrum measured 0.27 mm mrad in the experiment. The diagnosis of slice emittance facilitates local electron quality manipulation, which is important for the development of LWFA-based free electron lasers. The quasi-3D particle-in-cell simulations matched the experimental results and analysis well.

Under pressure to choose between the U.S. and China, Southeast Asian countries have adopted a hedging strategy: deepening economic relations with China while strengthening security cooperation with the U.S. How does the region's public view this strategy? With tensions rising in South China Sea territorial disputes, are more nationalistic individuals more likely to oppose hedging? Using an original public opinion survey conducted in the Philippines, we find that while an overwhelming majority of respondents were concerned about the territorial disputes, more nationalistic Filipinos were no more concerned than less nationalistic ones. Further, more nationalistic Filipinos were more likely to view economic relations with China as important for the Philippines and to approve of Duterte's China policy, which follows the logic of hedging. These surprising findings suggest that under the shadow of great-power competition, the link between domestic politics and foreign policy is nuanced in the Philippines, and Southeast Asia in general.

We report dispersion management based on a mismatched-grating compressor for a 100 PW level laser, which utilizes optical parametric chirped pulse amplification and also features large chirped pulse duration and an ultra-broadband spectrum. The numerical calculation indicates that amplified pulses with 4 ns chirped pulse duration and 210 nm spectral bandwidth can be directly compressed to sub-13 fs, which is close to the Fourier-transform limit (FTL). More importantly, the tolerances of the mismatched-grating compressor to the misalignment of the stretcher, the error of the desired grating groove density and the variation of material dispersion are comprehensively analyzed, which is crucially important for its practical application. The results demonstrate that good tolerances and near-FTL compressed pulses can be achieved simultaneously, just by keeping a balance between the residual second-, third- and fourth-order dispersions in the laser system. This work can offer a meaningful guideline for the design and construction of 100 PW level lasers.

Conditional value-at-risk (CVaR) and conditional expected shortfall (CES) are widely adopted risk measures which help monitor potential tail risk while adapting to evolving market information. In this paper, we propose an approach to constructing simultaneous confidence bands (SCBs) for tail risk as measured by CVaR and CES, with the confidence bands uniformly valid for a set of tail levels. We consider one-sided tail risk (downside or upside tail risk) as well as relative tail risk (the ratio of upside to downside tail risk). A general class of location-scale models with heavy-tailed innovations is employed to filter out the return dynamics. Then, CVaR and CES are estimated with the aid of extreme value theory. In the asymptotic theory, we consider two scenarios: (i) the extreme scenario that allows for extrapolation beyond the range of the available data and (ii) the intermediate scenario that works exclusively in the case where the available data are adequate relative to the tail level. For finite-sample implementation, we propose a novel bootstrap procedure to circumvent the slow convergence rates of the SCBs as well as infeasibility of approximating the limiting distributions. A series of Monte Carlo simulations confirm that our approach works well in finite samples.