Browse Articles
Discover research articles across all indexed journals
Analysis of lateral soil displacements induced by synchronous grouting in large-diameter shield tunnelling: a case study
Longitudinal analysis of the association between parity, mode of delivery and urinary incontinence in midlife using the SWAN cohort data
Aortic thrombi microstructure through contrast-enhanced X-ray microtomography
Siamese network with change awareness for surface defect segmentation in complex backgrounds
Abstract Despite the significant advancements made by deep visual networks in detecting surface defects at a regional level, the challenge of achieving high-quality pixel-wise defect detection persists due to the varied appearances of defects and the limited availability of data. To address the over-reliance on defect appearance and enhance the accuracy of defect segmentation, we proposed a Transformer-based Siamese network with change awareness, which formulates the defect segmentation under a complex background as change detection to mimic the human inspection process. Specifically, we introduced a novel multi-class balanced contrastive loss to guide the Transformer-based encoder, enabling it to encode diverse categories of defects as a unified, class-agnostic difference between defective and defect-free images. This difference is represented through a distance map, which is then skip-connected to the change-aware decoder, assisting in localizing pixel-wise defects. Additionally, we developed a synthetic dataset featuring multi-class liquid crystal display (LCD) defects set within a complex and disjointed background context. In evaluations using our proposed and two public datasets, our model outperforms leading semantic segmentation methods while maintaining a relatively compact model size. Furthermore, our model achieves a new state-of-the-art performance compared to semi-supervised approaches across various supervision settings. Our code and dataset are available at https://github.com/HATFormer/CADNet.
Strain-specific variations in the culture of chicken primordial germ cells
A hybrid evolution strategies algorithm for non-permutation flow shop scheduling problems
Abstract Flow shop scheduling has garnered significant attention from researchers over the past ten years, establishing itself as a prominent area of study within the field of scheduling. Nevertheless, there exists a paucity of research dedicated to addressing Non-Permutation Flow Shop Scheduling Problems. In this study, a Hybrid Evolution Strategies (HES) is suggested by combining the exploitation ability of Nawaz, Enscore, and Ham (NEH) Heuristic, the exploration ability of Improved Evolution Strategies (IES), and a Local Search Technique to minimize the makespan of NPFSSP. The primary solution is produced through the NEH Heuristic, serving as a foundational solution for the IES. The IES is applied in two stages, in the first stage it improves the permutation sequence found from the NEH heuristic. In the second stage of the IES, the permutation sequence on the first 40% of machines is fixed as found in the first stage. The sequence on the last 60% of machines is altered only so that the makespan is minimized and a good non-permutation sequence is found. Recombination and mutation are the main genetic operators in IES. For recombination in IES, 16 offspring are generated randomly from a single parent. The Quad swap mutation operator is employed in the IES to optimize the utilization of the solution space while minimizing computational time. To prevent trapping in local minima, a Local Search Technique is integrated into the IES algorithm, which guides solutions to less explored areas. Computational analyses indicate that HES exhibits superior performance regarding solution quality, computational efficiency, and robustness.
Defect detection in EBSM components through selective box fusion of modern object detection
Ulcerative colitis increases the risk of atrioventricular block: evidence from a Mendelian randomized analysis
A method for identifying common and unique issues in body in white dynamic stiffness based on modal contribution analysis
An ERP study on verb bias and thematic role assignment in standard Indonesian
Abstract This study investigates the effects of verb bias on sentence comprehension in Standard Indonesian (SI), a language where the passive construction is frequent and salient. We examined the hypothesis that lexical biases influence processing efforts. To disentangle effects of verb bias from structural effects, we held structure constant by using only passive sentences in our event-related potential (ERPs) experiment. Using an SI corpus, we identified verbs more frequently used in passive sentences (“passive-bias verbs”) and those more frequently used in active ones (“active-bias verbs”) to form passive sentences. Cluster-based permutation tests revealed significant differences between conditions. In the postverbal region (adverb), passive sentences with active-bias verbs elicited a broadly distributed negative shift that corresponded to the N280 distribution, a component associated with grammatical processing complexity. This suggests that active-bias verbs in passive contexts impose greater processing demands than their passive-bias counterparts, reflecting the integration of verb-specific preferences with sentence structure. These findings highlight how verb bias influences sentence comprehension and provide insights into cross-linguistic differences. In languages with predominantly active-bias verbs, using them in passive constructions may amplify processing costs. This study shows the importance of lexical biases and structural preferences in shaping sentence processing, contributing to a broader understanding of the interaction between lexical and syntactic factors in language comprehension.
Enhanced urease inhibitory activity of quercetin via conjugation with silver nanoparticles: synthesis, characterization, and DFT study
Combining semiparametric and machine learning approaches for short-term prediction of satellite clock bias
Patient-centered communication and its impact on minority health competence and outcomes in the United States
Association between vitamin A, E, and folate levels and risk of non-alcoholic fatty liver disease in adults with diabetes mellitus
Determination of the Median Effective Dose of Ciprofol Combined with Sufentanil in Inhibiting Tracheal Intubation Response in Female Patients
An improved scatter search algorithm for solving job shop scheduling problems with parallel batch processing machine
Depressive mixed state and anxious distress as risk factors for suicidal behavior during major depressive episodes
Abstract Accurately assessing and predicting suicidal behavior in patients with depression are challenging for researchers and clinicians. We examined various risk factors for suicidal behavior during major depressive episodes (MDE), especially focusing on depressive mixed state (DMX) and anxious distress (AD). We recruited 187 patients with MDE and divided them into two groups—with and without suicidal behavior—defined as the cut-off score of 1 or more on the suicidal behavior sub-item in the quick inventory of depressive symptomatology-self report. The presence of DMX was defined as a total score of 13 or more on the self-administered 8-item questionnaire for DMX. We used multivariate logistic regression analysis with the presence or absence of suicidal behavior as a dependent variable for investigating factors associated with suicidal behavior. The with suicidal behavior group was younger and indicated a greater proportion of past suicide attempts, AD, and DMX than the without suicidal behavior group. Logistic regression analysis revealed that AD ( P = 0.020) and DMX ( P = 0.018) were significantly associated with suicidal behavior. AD and DMX may promote suicidal behavior during MDE. These two psychopathological features should be carefully monitored and intensively treated for the prevention of suicide-related events.
Federated learning with integrated attention multiscale model for brain tumor segmentation
Abstract Brain tumors are an extremely deadly condition and the growth of abnormal cells that have formed inside the brain causes the illness. According to studies, Magnetic Resonance Imaging (MRI) is a fundamental imaging method that is frequently used in medical diagnostics to identify, treat, and routinely check for brain cancers. These images include extremely private and delicate details regarding the brain health of the individuals and it must be treated with much care to ensure anonymity of patients. However, traditional brain tumor segmentation techniques usually rely on centralized data storage and analysis, which might result in privacy issues and violations. Federated learning offers a solution by enabling the cooperative development of brain tumor segmentation models without necessitating the transfer of raw patient data to a centralized location. All the data are held securely within their institution. A Reinforcement Learning-based Federated Averaging (RL-FedAvg) model is proposed that fuses the Federated Averaging (FedAvg) model with Reinforcement Learning (RL). To optimize the global model for image segmentation jobs as well as to govern the consumption of client resources, the model dynamically updates client hyperparameters upon real-time performance feedback. A Double Attention-based Multiscale Dense-U-Net model, known as mixed-fed-UNet, is proposed in the work that uses the RL-FedAvg algorithm. The proposed technique achieves 98.24% accuracy and 93.28% dice coefficient on BraTs 2020 dataset. While comparing the developed model with the other existing methods, the proposed methodology shows better performance.