The proliferation of deepfake technology has precipitated a critical trust crisis in digital multimedia forensics, calling into question the reliability of existing detection systems. Current models predominantly rely on softmax-normalized probabilities, which exhibit heightened vulnerability to adversarial perturbations and out-of-distribution (OOD) samples. To address this deficiency and provide courts with quantifiably reliable forensic evidence, this paper proposes a Dual-Threshold Reliability Assessment framework (DTRA) grounded in class-conditional feature space analysis. The DTRA framework quantifies epistemic uncertainty through Mahalanobis distance-based inconsistency scoring computed from deep feature representations. Departing from conventional single-threshold paradigms, we independently calibrate optimal decision intervals for authentic and forged classes on a held-out calibration set. The interval optimization is formulated as a risk-adjusted utility maximization problem that trades off empirical precision against effective sample coverage. Specifically, an interval search algorithm identifies the most reliable subrange of inconsistency scores for each class via an odds ratio-weighted utility function, eschewing the restrictive assumption of a zero lower bound. DTRA serves as a conservative safeguard: when sample evidence is ambiguous, it abstains rather than forces a prediction. While this conservatism reduces coverage, the predictions it retains are significantly more reliable, thereby reducing the risk of high-confidence misjudgments.

ABSTRACT With the exponential growth of graph‐structured data, single‐machine efficient analysis has become increasingly impractical, making high‐performance distributed graph computing systems indispensable. The efficacy of these systems hinges critically on high‐quality graph partitioning. The edges of many graphs stemmed from real applications are uncertain, but many existing graph partitioning algorithms are only for deterministic graphs without considering uncertainty. This paper presents a novel partitioning algorithm, PAUG (Partitioning Algorithm for Uncertain Graphs), tailored for uncertain graphs. First, it formalizes the partitioning problem as an optimization task to minimize the cut‐edge ratio while balancing load. Second, it introduces probabilistic similarity to quantify vertex relationships under uncertainty. Finally, it details the PAUG algorithm which consists of initial partition phase and score‐function‐guided refinement strategy. Experimental results shows that PAUG achieves an average 23.2% reduction in cut‐edge ratio and a 26.2% improvement in load balance over state‐of‐the‐art algorithms.

Medical records are comprehensive repositories of patient health information and an essential tool for physicians to access medical histories. However, drafting medical records is a time-consuming process that contributes significantly to physician workload. Recent advances in Generative Artificial Intelligence (GAI) have shown strong potential in text summarization, yet most existing approaches rely on offline generation and singleturn interactions, failing to meet the real-time accuracy and user experience requirements of clinical practice. To address these limitations, we propose MedStructGen, a two-stage medical record generation framework that mirrors real-world clinical workflows. Stage 1 employs a simulation driven multiturn patient-physician dialogue model that adaptively refines its questioning strategy based on evolving symptom profiles, ensuring domain-complete diagnostic information capture. Stage 2 employs stage-specific fine-tuned LLMs with stepwise prompt engineering and entity alignment to generate EMRs that are compliant with standards and machine verifiable. This design not only improves contextual accuracy and completeness of information, but also achieves superior structural compliance, enabling seamless integration into Hospital Information Systems. Experiments on real-world hospital datasets demonstrate that our method achieves a 12.26% BLEU-4 improvement over one-stage baselines, with consistent gains in ROUGE and BERTScore. Located in a partner hospital, our system reduces physician documentation time by 1 to 1.5 hours per day, allowing more focus on patient care and personal well-being. The approach is generalizable and requires minimal customization for integration into other healthcare settings.

Passive microwave precipitation estimation from FengYun-3 (FY-3) polar-orbiting meteorological satellite plays an important role for global precipitation measurement. This paper carries out quality assessment of FY-3D microwave radiation imager (MWRI) instantaneous surface rain rate product (MRR) from 2018 to 2020 based on Global Precipitation Measurement (GPM) Dual-frequency Precipitation Radar (DPR) Level 2 product (2ADPR). Dichotomous and continuous statistics for quality assessment are analyzed in monthly scale. The results show that FY-3D MRR product, compared with 2ADPR, can identify the occurrence of rainfall well with a high Probability of detection and a low false alarm ratio, and can describe the rain intensity well with a high correlation coefficient and low relative difference. A case over northeast China region also shows a good performance of FY3D MRR product compared with ground Doppler radar data.

This study assimilated data from 9 Water Vapor(WV) channels of Metop-B IASI into the CMA-GFS 4DVAR system. Two sets of experiments, METOP-B-IR and CTRL, were designed, conducting an 18-day cycle assimilation test to evaluate the impact of assimilating MetOp-B IASI water vapor data on CMA-GFS model’s analysis and forecasting fields. By using ECMWF reanalysis data ERA5 as a reference, the study demonstrated that after assimilating 9 WV data, there was certain extent of improvement in the humidity and wind fields and led to slightly positive improvements in the forecasting skills.

Gestational diabetes mellitus (GDM) is a common complication of pregnancy, which has significant adverse effects on both the mother and fetus. The incidence of GDM is increasing globally, and early diagnosis is critical for timely treatment and reducing the risk of poor pregnancy outcomes. GDM is usually diagnosed and detected after 24 weeks of gestation, while complications due to GDM can occur much earlier. Copy number variations (CNVs) can be a possible biomarker for GDM diagnosis and screening in the early gestation stage. In this study, we proposed a machine-learning method to screen GDM in the early stage of gestation using cell-free DNA (cfDNA) sequencing data from maternal plasma. Five thousand and eighty-five patients from north regions of Mainland China, including 1942 GDM, were recruited. A non-overlapping sliding window method was applied for CNV coverage screening on low-coverage (~0.2×) sequencing data. The CNV coverage was fed to a convolutional neural network with attention architecture for the binary classification. The model achieved a classification accuracy of 88.14%, precision of 84.07%, recall of 93.04%, F1-score of 88.33% and AUC of 96.49%. The model identified 2190 genes associated with GDM, including DEFA1, DEFA3 and DEFB1. The enriched gene ontology (GO) terms and KEGG pathways showed that many identified genes are associated with diabetes-related pathways. Our study demonstrates the feasibility of using cfDNA sequencing data and machine-learning methods for early diagnosis of GDM, which may aid in early intervention and prevention of adverse pregnancy outcomes.

Multilayer clouds have a significant importance on cloud climate effects and remote sensing retrieval. In this study, a multilayer cloud detection algorithm is developed for the Advanced Geostationary Radiation Imager (AGRI) onboard the FY-4A geostationary satellite. The algorithm is based on the basic physical assumptions that are also employed for Moderate Resolution Imaging Spectroradiometer (MODIS) and Visible Infrared Imager Radiometer Suite (VIIRS) to identify ice overlaying water multilayer clouds. Synchronous observation of Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) has been collected and acknowledged as a reliable reference dataset for determination of thresholds. The algorithm used the long-wave infrared bands (8.5 and$10.8~\mu \text{m}$) to determine the phase of the upper layer cloud. Then, the difference between solar reflectance band pairs (1.375 and$1.61~\mu \text{m}$) is used to identify ice overlayer water multilayer clouds when the upper layer is ice cloud. When the upper layer cloud is water, the infrared band ($7.1~\mu \text{m}$) is applied to find misclassified multilayer clouds. The algorithm demonstrates a notable improvement of approximately 0.146 in the probability of detection (POD) compared to MODIS while using CALIOP products as a reference, specifically for cases when the cloud optical depth (COD) surpasses 4. Nevertheless, it does result in a slightly elevated false alarm rate (FAR), around 0.042. In the future, it is necessary to conduct a more comprehensive validation of the algorithm, with particular emphasis on its limits in scenarios where the upper cloud layer is too thin (thick).

By leveraging deep neural networks, recent face swapping techniques have performed admirably in generating faces that maintain consistent identities. Nevertheless, while these methods accurately transfer source identities, they often struggle to preserve important attributes (such as head poses, expressions, and gaze directions) in the target faces. As a consequence, the current research in this domain has not resulted in satisfactory performance. In this paper, we propose an efficient attribute-preserving framework, called AP-Swap, for short, for face swapping. Our approach incorporates two innovative modules designed specifically to preserve critical facial attributes. First, we propose a global residual attribute-preserving encoder (GRAPE), which adaptively extracts globally complete attribute features from target faces. Second, in addition to the regular network streams for the source and target facial images, we introduce a network stream that takes into account the facial landmarks of the target faces. This additional stream enables our landmark-guided feature entanglement module (LFEM), which efficiently preserves fine-grained facial attributes by conducting a landmark-based attribute-preserving (LBAP) operation. Through extensive quantitative and qualitative experiments, we demonstrate the superiority of AP-Swap over other state-ofthe-art methods in terms of facial attribute preservation and model efficiency, along with satisfactory identity consistency performance

In this paper, a novel denoising method for electrocardiogram (ECG) signal is proposed to improve performance and availability under multiple noise cases. The method is based on the framework of conditional generative adversarial network (CGAN), and we improved the CGAN framework for ECG denoising. The proposed framework consists of two networks: a generator that is composed of the optimized convolutional auto-encoder (CAE) and a discriminator that is composed of four convolution layers and one full connection layer. As the convolutional layers of CAE can preserve spatial locality and the neighborhood relations in the latent higher-level feature representations of ECG signal, and the skip connection facilitates the gradient propagation in the denoising training process, the trained denoising model has good performance and generalization ability. The extensive experimental results on MIT-BIH databases show that for single noise and mixed noises, the average signal-to-noise ratio (SNR) of denoised ECG signal is above 39 dB, and it is better than that of the state-of-the-art methods. Furthermore, the denoised classification results of four cardiac diseases show that the average accuracy increased above 32 % under multiple noises under SNR=0 dB. So, the proposed method can remove noise effectively as well as keep the details of the features of ECG signals.

Data assimilation of satellite microwave sounders are very important for numerical weather prediction. Fengyun-3 (FY-3) polar satellites carry two such sounders: MicroWave Temperature Sounder (MWTS) and MicroWave Humidity Sounder (MWHS). The microwave observations should be quality-controlled before assimilation and rain check is one of the crucial steps. Microwave Radiation Imager (MWRI), on board FY-3 too, can provide useful rain information due to more sensitive to precipitation than sounders. A remapping methodology is described in this paper to make full use of data from sounders and imager simultaneously. Observations and products from MWRI can thus be remapped to field of view of sounders. Comprehensive information obtained in this way can be used to not only assimilation in numerical weather prediction models but also satellite observations quality monitoring.

The Tibetan Plateau has a significant impact on the precipitation in the lower reaches. Forecasted precipitation with assimilated satellite data in plateau is often closer to the actual situations. Using Gridpoint Statistical Interpolation (GSI) system and The Weather Research and Forecasting Model (WRF), three experiments were designed to evaluate the impacts of MHS data assimilation of Tibetan Plateau on lower reaches rainfall forecasts. The results showed that the assimilation of satellite MHS radiance can improve the accuracy of the prediction of precipitation intensity and precipitation zone in the lower reaches. A preliminary analysis of the reasons showed that the MHS data assimilation affected the initial water vapor fields, such that the model had better humidity conditions at the initial stage. The assimilation of conventional wind data increased the wind speed over the plateau. The interaction of water vapor and wind fields increased the accuracy of precipitation forecast in the study case.

A cross-sensor calibration technique is developed and applied to improve upon the prelaunch radiance calibration and characterization for the Total Ozone Unit (TOU) onboard the FengYun-3/A satellite. The Level 3 products from the National Aeronautics and Space Administration Ozone Monitoring Instrument (OMI) onboard the Earth Observing System Aura are used as input to a radiative transfer model to predict the TOU radiances and characterize the biases for the measurements over the Pacific Ocean in low- and midlatitudes. The coefficients are derived from a regression algorithm to adjust the TOU radiances. It is shown that, after the measurement bias correction, the biases between the retrieved total column ozone products from the TOU with those from the OMI Total Ozone Mapping Spectrometer (TOMS)-Version 8 products and those from a set of ground-based station measurements are 3 % and 5% , respectively. The variations in the estimated total ozone amounts from the TOU are consistent with those derived from Solar Backscatter Ultraviolet Radiometer instruments and OMI for a period from January 2010 to February 2011.