With the rapid growth of video content on social media, video summarization has become a crucial task in multimedia processing. However, existing methods face challenges in capturing global dependencies in video content and accommodating multimodal user customization. Moreover, temporal proximity between video frames does not always correspond to semantic proximity. To tackle these challenges, we propose a novel Language-guided Graph Representation Learning Network (LGRLN) for video summarization. Specifically, we introduce a video graph generator that converts video frames into a structured graph to preserve temporal order and contextual dependencies. By constructing forward, backward and undirected graphs, the video graph generator effectively preserves the sequentiality and contextual relationships of video content. We designed an intra-graph relational reasoning module with a dual-threshold graph convolution mechanism, which distinguishes semantically relevant frames from irrelevant ones between nodes. Additionally, our proposed language-guided cross-modal embedding module generates video summaries with specific textual descriptions. We model the summary generation output as a mixture of Bernoulli distribution and solve it with the EM algorithm. Experimental results show that our method outperforms existing approaches across multiple benchmarks. Moreover, we proposed LGRLN reduces inference time and model parameters by 87.8% and 91.7%, respectively.

Benzene exposure biomarkers are associated with recently smoking tobacco and pumping gasoline in the U.S. population aged 12 and over: NHANES 2017-March 2020.

The flexibility of finite mixture models makes them suitable candidates for analyzing survival data with complex, multimodal distributions. Such data is often available if the event of interest occurs due to multiple failure modes. Here, we explore the modeling of competing risks time-to-event data with covariates in the presence of long-term survivors in the population using finite mixture models. The mixture cure rate model is used to describe the uncertainty in the population, where the susceptible part of the population is modeled using a finite mixture of Weibull distributions with different shape and scale parameters. Moreover, if information on covariates is available, the cure rate may be modeled using a binary regression model on the covariates. Here, we use the logistic function to relate covariates to the cure rate. The distribution corresponding to the susceptible part may also depend on covariates. To explore such dependency, we model the scale parameter of the Weibull distribution using covariates. Then, we discuss the classical parametric inference for the constructed model based on random and non-informative right-censored competing risks time-to-event data. An efficient method based on the expectation-maximization algorithm is proposed to estimate model parameters, thereby avoiding the complexity of directly maximizing the likelihood function. Additionally, a method for constructing confidence intervals for all model parameters is addressed. A simulation study is performed in the presence of two competing causes to investigate the finite sample properties of the proposed estimation methodologies. Finally, the methods are illustrated by analyzing a real data set on malignant melanoma cancer. Predicting the conditional survival function of an alive patient is of natural interest to an experimenter or medical researcher. A method for estimating such a conditional survival probability is also discussed.

Active inference, computational phenomenology, and advanced meditation: Toward the formalization of the experience of meditation.

Cost-effectiveness of Preventive Transarterial Embolization for Splenic Artery Aneurysm Below the Guideline-Recommended Size Threshold: A Japanese Claims-Based Study.

Automatic seizure detection holds significant importance for epilepsy diagnosis and treatment. Convolutional neural networks (CNNs) have shown immense potential in seizure detection. Though traditional CNN-based seizure detection models have achieved significant advancements, they often suffer from excessive parameters and limited interpretability, thus hindering their reliability and practical deployment on edge computing devices. Therefore, this study introduces an innovative Morlet convolutional neural network (Morlet-CNN) framework with its effectiveness demonstrated in seizure detection tasks. Unlike traditional CNNs, the convolutional kernels in the Morlet-CNN contain only two learnable parameters, allowing for a lightweight architecture. Additionally, we propose a frequency-domain-response-based kernel pruning algorithm for Morlet-CNN and implement an INT8 quantization algorithm by incorporating Kullback-Leibler (KL) divergence calibration with a Morlet lookup table (LUT). With the pruning and quantization algorithms, the model's parameter scale achieves over 90% reduction while maintaining minimal accuracy loss. Furthermore, the model exhibits enhanced interpretability from a signal processing perspective, distinguishing it from many previous CNN models. Extensive experimental validation on the Bonn and CHB-MIT datasets confirms the Morlet-CNN model's efficacy with a compact Kilobyte (KB)-level model size, making it highly suitable for real-world applications.

We propose a method which, given a sequence of stereo foggy images, estimates the parameters of a fog model and updates them dynamically. In contrast with previous approaches, which estimate the parameters sequentially and thus are prone to error propagation, our algorithm estimates all the parameters simultaneously by solving a novel optimisation problem. By assuming that fog is only locally homogeneous, our method effectively handles real-world fog, which is often globally inhomogeneous. The proposed algorithm can be easily used as an add-on module in existing visual Simultaneous Localisation and Mapping (SLAM) or odometry systems in the presence of fog. In order to assess our method, we also created a new dataset, the Stereo Driving In Real Fog (SDIRF), consisting of high-quality, consecutive stereo frames of real, foggy road scenes under a variety of visibility conditions, totalling over 40 minutes and 34 k frames. As a first-of-its-kind, SDIRF contains the camera's photometric parameters calibrated in a lab environment, which is a prerequisite for correctly applying the atmospheric scattering model to foggy images. The dataset also includes the counterpart clear data of the same routes recorded in overcast weather, which is useful for companion work in image defogging and depth reconstruction. We conducted extensive experiments using both synthetic foggy data and real foggy sequences from SDIRF to demonstrate the superiority of the proposed algorithm over prior methods. Our method not only produces the most accurate estimates on synthetic data, but also adapts better to real fog.

Imposing accurate outflow boundary conditions remains a significant challenge in 3D computational fluid dynamics simulations of patient-specific cerebral blood flow. Widely used Windkessel models often rely solely on geometric factors, such as outlet numbers and diameters, leading to inaccuracies caused by image quality limitations and simplified vessel representations. This preliminary study proposes a novel functional region-based approach to enhance the accuracy of cerebral blood flow simulations. Cerebral vessels were divided into functional regions by combining population-based cerebral blood flow distributions with patient-specific arterial geometries from medical images. Within each functional region, parameters of Windkessel models for individual outlets are calculated based on their corresponding diameters/areas, accounting for both functional and geometric characteristics. Validation was conducted on a single subject using clinical Transcranial Doppler ultrasound data, with comparisons made to a conventional area-based approach. The functional region-based approach demonstrated better alignment with clinical measurements, outperforming the area-based method in velocity profiles at 5 of 7 monitored locations. It also provided closer agreement with measured blood flow distribution, with maximum percentage differences of -4.5%. By integrating vascular geometry and functional perfusion data, the proposed approach provides a physiologically informed strategy for setting outlet boundary conditions in cerebral blood flow simulations. Although demonstrated in a single-subject case, this approach shows potential to improve patient-specific simulation reliability by reducing errors caused by imaging artifacts and geometric simplifications, offering value for future clinical and research applications.

Daily global transpiration estimation (2001–2018) by integrating satellite solar-induced fluorescence and spatially heterogeneous slope parameter in a conductance-photosynthesis model

To investigate the effects of combined exercise training associated with enalapril maleate on blood pressure variability (BPV) and renal morphofunctional, inflammatory and oxidative stress parameters in an experimental model of arterial hypertension. Male spontaneously hypertensive rats (SHR) were randomly allocated into sedentary placebo (SP), trained placebo (TP), sedentary enalapril (SE) or trained enalapril (TE). Both enalapril treatment (3 mg/kg) and combined exercise training (3 days/week) were performed for 8 weeks. Blood pressure (BP) was recorded intra-arterially for BPV analysis. Renal function, morphology, inflammation and oxidative stress were assessed. Combined exercise training alone (TP group) did not alter systolic BP. However, TP group showed lower media/lumen ratio of interlobular arteries and NADPH oxidase activity, as well as higher interleukin (IL)-10 and superoxide dismutase activity in renal tissue compared to the SP group. In addition to similar benefits induced by exercise training alone, the combination of approaches (TE group) resulted in lower vascular sympathetic modulation (TE: 10.6 ± 1.7 vs. SP: 22.0 ± 3.1 mmHg 2 ), higher creatinine clearance, lower NADPH oxidase activity, lower areas with severe tubulointerstitial fibrosis (injury range 51-100%, TE: 10.0 ± 0.2 vs. SP: 27.5 ± 0.1, TP: 22.5 ± 0.1 and SE: 22.5 ± 0.1%), as well as a lower media/lumen ratio. Positive correlations were obtained between vascular sympathetic modulation with SBP ( r = 0.61), media/lumen ratio ( r = 0.74) and renal tubulointerstitial fibrosis ( r = 0.69). The combination of exercise training with enalapril provided additional renal morphofunctional benefits, which may result from interactions involving BPV, inflammation, and oxidative stress, and could contribute to the observed renal improvements. Our findings also suggest that BPV may play a role in hypertension-related renal changes and that combining pharmacological and nonpharmacological therapies might offer effective strategies to reduce residual cardiovascular risk in arterial hypertension.

Likelihood ratio estimation of partial Y-STR profile matches using discrete Laplace models and marginalisation.

Recognition of immunogenomic signature and prognostic value of the subtype of epithelial-mesenchymal transition in breast cancer.

Calibration and experimental verification of tillage model parameters for yellow-brown loam soil in paddy fields based on the discrete element method

Output and transfer characteristics prediction of GaN HEMT via neural-network-based compact model parameters extraction

Automatic colony cell counting method of phytoplankton in microscopic images with multi-task learning.

A multiobjective optimization approach to data assimilation for complex biological systems with sparse data.

In situ amidation of NH3 in biogas on sludge-derived carbon to promote coexisting CO2 adsorption.

The effect of base-rate priors on decision-making and confidence in healthy aging.

Sleep homeostasis describes the accumulation of sleep pressure during wakefulness and its dissipation during sleep. Electroencephalographic delta power (SWA, [0.5-4Hz]) is a known marker of sleep pressure. Dysregulation of this process has been proposed in idiopathic hypersomnia (IH), a debilitating disorder characterized by severe daytime sleepiness and excessive sleep duration. We investigated SWA regulation in IH with (IH-LST) and without (IH-nLST) long-sleep time and healthy participants in a controlled ad libitum sleep protocol. We recorded SWA in 208 IH participants and 25 controls during a 32-hour continuous polysomnography without circadian cues. We characterized sleep architecture and SWA as functions of sleep and prior wake duration, to capture dissipation and accumulation of sleep pressure. We then applied an established model of sleep pressure fitted individually. Model parameters were compared across groups, and simulations were performed to predict and compare sleep pressure during extended sleep and wakefulness. In all IH participants (88% IH-LST, 12% IH-nLST), SWA persisted during extended sleep, while wakefulness led to greater SWA rebounds only in IH-LST; indicating dysregulation of accumulation and dissipation. In the mathematical approach, sleep pressure failed to dissipate after nine hours of sleep in all IH participants, and accumulated faster during the first four hours of wakefulness in the IH-LST subgroup. These effects were mediated by a stronger accumulation during both early wakefulness and late REM sleep. Our findings highlight a dysregulation of sleep homeostasis in idiopathic hypersomnia and support the reintroduction of the distinction between the phenotypes with and without long sleep time.

Increasing charging demand with the widespread use of electric vehicles leads to negative effects such as load imbalance, sudden load changes, harmonics and voltage fluctuations in the electricity distribution network. Furthermore, irregular charging demand negatively impacts electric vehicle user comfort and traffic management. This study presents a dynamic pricing-based energy management model developed for use in urban electric vehicle charging infrastructures to address these challenges. The proposed model considers price not only as an economic output but also as a control variable that manages grid load balance. There are four input parameters (traffic, station occupancy rate, location and state of charge) in the pricing model and these parameters are dynamically updated at each iteration. The model was developed in MATLAB environment and was employed real-time traffic data obtained through the Google Maps API. The model tested for ten iterations. The results show that the pricing model prioritizes low charge levels vehicles. But the model maintaining balanced grid load simultaneously. Furthermore, price output increases high occupancy rates charging stations in order to encourage users to choose stations with lower occupancy rates. Results of this study demonstrates that pricing mechanism can be used as a decision variable both economic reasons and system efficiency. In future works, the model might be extended with artificial intelligence and optimization-based methods. Pricing model serves as a potential solution to challenges in energy and transportation networks with the help of test systems.