Feng Chia University

Deepfake technology refers to the use of deep learning to generate fabricated audiovisual content. It can be utilized in virtual reality and movie production. However, if used for malicious purposes, it can impact individuals and society and pose national security issues. Governments worldwide have introduced policies and legislation to address these issues. Detecting deepfake videos is thus critically important. This study analyzes methods of generating and detecting deepfake videos. We used frameworks such as DeepFaceLab and FakeApp to produce deepfake videos. Employing the Xception depthwise separable convolution network model as the foundation, we adopted the FakeVideoForensics method for continuous video analysis, the DeepFakes_FacialRegions method for specific facial feature analysis, and the Improved Xception method, which showed better performance. We have developed an enhanced depthwise separable convolution and facial feature extraction method for deepfake video detection, named Improved Xception Feature Fake Video Detection (IXFFVD). This method was trained using UADFV, FaceForensics++, Celeb-DF, and DFDC datasets, and it was used to detect face-swapped deepfake videos created by frameworks such as DeepFaceLab and FakeApp. Experimental results show that IXFFVD outperforms the compared methods. Although the detection efficacy of this method could be further enhanced, future work will focus on incorporating spatiotemporal detection criteria to improve the detection of deepfake videos created from different datasets.

Abstract The cone penetration test (CPT) is essential for soil classification and pile bearing capacity assessment. In practice, engineers simplify stratigraphic profiles into representative layers, but this process often depends on subjective judgment, introducing uncertainty. This study proposes an automated CPT-based soil profile simplification framework that preserves key stratigraphic characteristics while minimizing subjectivity. The method determines representative layer thickness based on averaged CPT-derived data and applies systematic merging criteria. Over 50 stratigraphic profiles from Taiwan and Germany were analyzed and compared with results from commercial software. The results show that the axial pile capacity deviations between the simplified and original soil profiles depend strongly on both the T/D ratio and the regional soil conditions. For Taiwanese offshore silty soils, the errors increase with larger T/D values, ranging from 4–42% at T/D=0.25 to as high as 82% at T/D=1.0. In contrast, for German offshore sandy soils, the deviations remain below 20% across all methods and thicknesses. These findings indicate that the proposed simplification approach provides a practical balance between computational efficiency and accuracy, but the acceptable T/D range should be site- and method-dependent.

IntroductionCollision tumors are extremely rare in the sellar region, less than 1 to 2% of sellar masses. Intracranial dermoid cyst is a rare, benign, slow-growing lesion accounting for 0.04 to 0.7% of all intracranial tumors. Sellar dermoid cysts are extremely rare; from 1976 to 2024, there were only 59 records of sellar dermoid cyst published in the English language. Coexisting sellar dermoid cyst and pituitary adenoma have not been reported in the literature.Case PresentationA 51-year-old man having medical history of left traumatic optic neuropathy and acute lymphoblastic leukemia posttreatment with complete remission in 2018 suffered from intermittent dizziness and occasional unsteady gait with deviation to the left side for more than 6 months prior to visiting our hospital. His brain magnetic resonance imaging disclosed a mass lesion in the sellar region with suprasellar extension, and his preoperative hormone study showed hypopituitarism. After a thorough preoperative evaluation, he underwent an endoscopic endonasal transsphenoidal approach with the removal of the lesion and skull base reconstruction. The lesion was pathologically diagnosed to be a dermoid cyst with coexisting pituitary adenoma/pituitary neuroendocrine tumor.ConclusionTo our best knowledge, this is probably the first report of sellar dermoid cyst with coexisting pituitary adenoma/pituitary neuroendocrine tumor.

This article investigates the deflection behavior of Euler-Bernoulli beams subjected to axial compression and resting on a nonlinear elastic foundation. The Modified Adomian Decomposition Method (MADM) is employed to predict the beam deflection under various loading and foundation conditions. By adopting an initial polynomial ansatz, MADM effectively optimizes the construction of Adomian polynomials, resulting in rapid convergence and an accurate series solution. The accuracy and reliability of the proposed method are examined through two illustrative cases. In the first case, the formulation is verified against an analytically tractable benchmark problem, while in the second case, the MADM results are compared with previously published solutions. The comparative analysis demonstrates excellent agreement with existing studies, confirming the validity of the proposed approach. Overall, the results indicate that MADM provides a stable and efficient analytical framework for modeling the coupled effects of axial compression and nonlinear elastic foundations in Euler-Bernoulli beams.

Fuzzy-set qualitative comparative analysis (fsQCA) has gained widespread popularity in social science research. However, incorporating the concept of growth—common in business and economic studies—into fsQCA remains unintuitive and methodologically challenging. This study proposes a new approach to systematically integrate growth into fsQCA applications. The method addresses three key challenges: the naming of antecedents and outcomes, the calibration of variables, and the interpretation of results. An empirical analysis demonstrates that the proposed approach offers a more generalized and robust framework for handling both positive and negative growth values. By applying the method, antecedents and outcomes are appropriately named and calibrated, and the resulting solutions are more interpretable and self-explanatory. Ultimately, the method enhances the clarity and accuracy of fsQCA outcomes, reducing the risk of misinterpretation.

As network communication technologies rapidly advance, ransomware has emerged as a significant cybersecurity threat that organizations cannot ignore. Static analysis enables rapid identification of ransomware by examining file structure and code characteristics before execution. However, existing classifiers are predominantly designed under the closed-set assumption, causing them to misclassify novel variants into known families. Furthermore, ransomware datasets typically exhibit long-tailed distributions with emerging families having very few available samples, making it difficult for models to learn discriminative features. To address these challenges, we propose Few-Shot Open-Set Ransomware Detection through Meta-learning and Energy-based Modeling (MEM), a unified open-set recognition framework based on static analysis of Portable Executable features. By integrating Model-agnostic Meta-learning (MAML), the model rapidly adapts to new families with limited samples. The Energy Function quantifies the confidence of predictions in distinguishing between known samples and unknown ones, while Focal Loss dynamically adjusts sample weights to reduce bias introduced by imbalanced distributions. The experimental results demonstrate that MEM achieves higher classification accuracy and better rejection performance of unknown samples than existing open-set recognition methods.