Search Results (34,824)

Facial beauty prediction (FBP) is a significant research direction in the field of computer vision; however, the performance of models developed for this task is often constrained due to the scarcity of high-quality annotated data. Generative adversarial networks (GANs) are efficient image generation networks that are capable of rapidly generating facial images. This study proposes an FBP method—named FBP-GAN—which aims to address this shortage of data by generating high-quality synthetic facial images. First, we construct a facial image generation network based on StyleGAN2-ADA to generate diverse and realistic facial images. Second, we combine transfer learning and data augmentation techniques to utilize the synthesized images for training set augmentation while optimizing the category distribution to enhance the generalization ability and prediction accuracy of the classification network. The experimental results demonstrate that, when using MobileViT or ResNeXt as the classification network, our proposed approach achieves prediction accuracies of 76.38% and 77.94% on the SCUT-FBP5500 dataset, respectively, representing improvements of 0.55% and 1.65% over the baseline models’ 75.83% and 76.29%. The proposed approach effectively improves the accuracy of FBP under data-scarce scenarios and opens new avenues for the application of GANs in computer vision tasks. Full article

This study presents an edge-based intrusion detection methodology designed to enhance cybersecurity in Internet of Things environments, which remain highly vulnerable to complex attacks. The approach employs an Auxiliary Classifier Generative Adversarial Network capable of classifying network traffic in real-time while simultaneously generating high-fidelity synthetic data within a unified framework. The model is implemented in TensorFlow and deployed on the energy-efficient NVIDIA Jetson Orin Nano, demonstrating the feasibility of executing advanced deep learning models at the edge. Training is conducted on network traffic collected from diverse IoT devices, with preprocessing focused on TCP-based threats. The integration of an auxiliary classifier enables the generation of labeled synthetic samples that mitigate data scarcity and improve supervised learning under imbalanced conditions. Experimental results demonstrate strong detection performance, achieving a precision of 0.89 and a recall of 0.97 using the standard 0.5 decision threshold inherent to the sigmoid-based binary classifier, indicating an effective balance between intrusion detection capability and false-positive reduction, which is critical for reliable operation in IoT scenarios. The generative component enhances data augmentation, robustness, and generalization. These results show that combining generative adversarial learning with edge computing provides a scalable and effective approach for IoT security. Future work will focus on stabilizing training procedures and refining hyperparameters to improve detection performance while maintaining high precision. Full article

This article presents an improved formulation of Newton’s law of cooling using the conformable fractional derivative to model long-term thermal behavior more accurately. A key feature of our approach is the use of the fractional time variable $t^{\gamma }$, which introduces a simple scaling symmetry: the structure of the model remains unchanged even when time is proportionally stretched or compressed. This symmetry-based property provides additional flexibility compared to the classical formulation and enables the derivation of analytical solutions under both constant and non-constant ambient temperature. In particular, we incorporate sinusoidal models for ambient temperature to capture realistic environmental fluctuations over extended periods. Experimental measurements confirm that the conformable model achieves significantly better accuracy than traditional integer-order models. These results highlight the relevance of symmetry and fractional calculus in describing physical processes and demonstrate the potential of conformable methods for improving long-term thermal predictions. Full article

The global shortage of ${}^3$He resources has created an urgent need for alternative neutron detection technologies in applications such as national security, neutron scattering, and nuclear energy. This study designed and developed a zero-dimensional planar high-efficiency thermal neutron detector based on a boron-lined multi-wire proportional chamber (MWPC) employing two distinct efficiency-enhancement approaches: a multilayer structure and grazing-incidence geometry. For ease of use, a sealed detector has been developed, eliminating the need for gas cylinders. Geant4 simulations were utilized to optimize the ${\mathrm{B}}_4\mathrm{C}$ thickness of conversion layer and evaluate $\gamma$-ray sensitivity. Prototype detectors were fabricated and experimentally validated at the 20th beamline (BL20) of China Spallation Neutron Source (CSNS). Simulation results indicate that the optimal ${\mathrm{B}}_4\mathrm{C}$ thickness varies with layer count and neutron wavelength, measuring approximately 2.0 µm at 1.8 Å and 1.5 µm at 4 Å for a 10-layer structure, with $\gamma$-ray sensitivity below $5\times {10}^{-6}$. Experimental measurements demonstrate that a five-layer detector achieved neutron detection efficiencies of 28.0 ± 1.5% at 4.78 Å and 17.8 ± 1.8% at 2.87 Å, while a two-layer detector at 11.5${}^{\circ }$ incidence attained 19.2% and 11.7%. This research lays the groundwork for developing large-area, high-efficiency, position-sensitive neutron detectors Full article

Efficient autonomous exploration in unknown environments is a core challenge for Unmanned Aerial Vehicle (UAV) applications in unstructured settings. The primary challenges are exploration speed, coverage efficiency, and the autonomous, efficient, and obstacle-/threat-avoiding global guidance of UAV under local observational information. This paper proposes an autonomous exploration method driven by simultaneous incremental map prediction and the fusion of global frontier information to enhance the exploration efficiency of UAVs in unknown unstructured environments. Based on generative deep learning, we introduce an incremental map prediction method for 3D unstructured mountainous terrain, enabling the simultaneous acquisition of map predictions and their uncertainty estimates. Map prediction and trajectory planning are conducted concurrently: by utilizing the simultaneously predicted 3D map and its confidence (i.e., the uncertainty estimates), an overlap analysis is conducted between the flyable areas in the predicted map and the high-confidence regions. Dynamic guidance subspaces are generated by extracting global frontier points, within which shortest-time optimization is adopted for trajectory planning to maximize information gain and coverage per step. Experimental results demonstrate that compared to classical methods, our proposed approach achieves significant performance improvements in key metrics, including map coverage rate, total exploration time, and average path length. Full article

This study proposes an approach for detecting alcohol intoxication from speech based on a combination of audio segmentation and a hybrid neural network architecture that integrates convolution neural network (CNN) and long-short term memory (LSTM) layers. The proposed design enables effective modeling of both local spectral patterns and long-term temporal dependencies in speech signals. By operating on relatively long audio segments, the approach allows the simultaneous analysis of complex speech constructions and pause patterns, which are known to be sensitive to alcohol-induced speech impairments. Each audio signal was divided into two equal-duration segments that are processed sequentially by the model, which helps reduce the impact of asymmetrical distribution of intoxication-related speech artifacts. The approach was evaluated using the GradusSpeech-v1 corpus, which contains more than 1300 recordings of Russian tongue twisters collected from 31 speakers under controlled conditions in both sober and intoxicated states. Experimental results demonstrate that the proposed method achieves high performance. When full recordings are analyzed using median aggregation of segment-level predictions, the model reaches Accuracy, Recall, and F1-score values close to 0.93, indicating the effectiveness of the approach for alcohol intoxication detection in speech. Full article

This study demonstrates a Mach–Zehnder-type internal-state atom interferometer in a sodium F = 1 spinor Bose–Einstein condensate (BEC), which is realized by applying a three-pulse radio-frequency sequence ($\pi /2$–$\pi$–$\pi /2$) to manipulate the two magnetic sublevels $|1,-1⟩$ and $|1,0⟩$. Phase-scanning experiments show that the visibility remains at a high level across all three pulse stages ($V>0.77$). In the hold-time scanning measurements, the visibility decays exponentially with hold time, yet the system maintains good coherence. This work establishes a foundation for precision measurements based on internal-state atom interferometers, as the approach simplifies the experimental apparatus while maintaining good quantum coherence and high-contrast interference fringes. Full article

In the information age, Internet of Things (IoT) devices are more susceptible to intrusion due to today’s complex network attack methods. Therefore, accurately detecting evolving network attacks from complex and ever-changing IoT environments has become a key research goal in the current intrusion detection field. Due to the spatial and temporal characteristics of IoT data, this paper proposes a Spatiotemporal Feature Weighted Fusion Approach Combining Gating Attention Transformation (STWGA). STWGA consists of three parts, namely spatial feature learning, the gated attention transformer, and the temporal feature learning module. It integrates improved convolutional neural networks (CNN), batch normalization, and Bidirectional Long Short-Term Memory Network (Bi-LSTM) to fully learn the deep spatial and temporal features of the data, achieving the goal of global deep spatiotemporal feature extraction. The gated attention transformer introduces an attention mechanism. In addition, an additional control mechanism is introduced in the self-attention module to more effectively improve detection accuracy. Finally, the experimental results show that STWGA has better spatiotemporal feature extraction ability and can effectively improve the intrusion detection effect of anomalies. Full article

Accurate genotyping of small insertions and deletions (InDels; <5 bp) remains technically challenging in routine molecular breeding, largely due to the limited resolution of agarose gel electrophoresis and the labor-intensive nature of polyacrylamide-based assays. Here, we present the Tri-Primer Amplification Refractory Mutation System (TP-ARMS), a simple and cost-effective PCR-based strategy that enables high-resolution genotyping of small InDels using standard agarose gels. The TP-ARMS employs a universal reverse primer in combination with two allele-specific forward primers targeting insertion and deletion alleles, respectively. This design allows clear discrimination of homozygous and heterozygous genotypes using a two-tube PCR workflow. The method showed complete concordance with Sanger sequencing in detecting 1–5 bp InDels across multiple crop species, including rice (Oryza sativa) and quinoa (Chenopodium quinoa). In addition, using a TP-ARMS reduced experimental time by approximately 90% compared with PAGE-based approaches and avoided the high equipment and DNA quality requirements of fluorescence-based assays. The practical applicability of the TP-ARMS was demonstrated in breeding populations, including efficient genotyping of a 3-bp InDel in OsNRAMP5 associated with cadmium accumulation and a 6-bp promoter InDel in OsSPL10 underlying natural variation in rice trichome density across 370 accessions. Collectively, the TP-ARMS provides a robust, scalable, and low-cost solution for precise small InDel genotyping, with broad applicability in marker-assisted breeding and functional genetic studies. Full article

Bolt connections are widely used in engineering structures but are susceptible to loosening during operation, which can result in significant safety concerns. Consequently, reliable bolt-loosening detection is of paramount importance. Conventional detection methodologies frequently exhibit deficiencies, including reduced efficiency, constrained accuracy, and the requirement for contact sensors. To overcome these limitations, this study proposes a novel non-contact approach for bolt preload monitoring based on Digital Image Correlation (DIC). In this method, an industrial camera captures speckle images of the bolt head before and after deformation, thereby enabling measurement of the surface strain. The DIC technique is employed to calculate the strain field on the bolt head surface, which exhibits a linear relationship with the bolt preload. The proposed method utilizes strain field tracking to facilitate effective and precise monitoring of bolt preload. Experimental results demonstrate that the method provides a precise, efficient, and user-friendly solution for bolt preload monitoring, showing great potential for applications in structural health monitoring. Full article

Triple-negative breast cancer (TNBC) is an aggressive malignancy that disproportionately affects young women. The integration of immune checkpoint inhibitors (ICIs) has significantly improved outcomes in both early-stage and metastatic TNBC, shifting attention toward long-term survivorship issues, particularly endocrine function and fertility. However, the reproductive safety profile of ICIs remains insufficiently characterized. This narrative review synthesizes current preclinical and clinical evidence on ICI-associated reproductive toxicity, focusing on both direct immune-mediated gonadal injury and indirect disruption of the hypothalamic–pituitary–gonadal axis. Experimental models consistently demonstrate immune cell infiltration of ovarian and testicular tissue, cytokine-driven inflammatory cascades, follicular atresia, impaired spermatogenesis, and altered steroidogenesis following PD-1/PD-L1 and CTLA-4 blockade. Emerging clinical data report cases of immune-related orchitis, azoospermia, testosterone deficiency, diminished ovarian reserve, and premature ovarian insufficiency. Secondary hypogonadism due to immune-mediated hypophysitis represents an additional and frequently underdiagnosed mechanism. We further discuss the oncofertility challenges faced by young patients with TNBC treated with chemoimmunotherapy, emphasizing the uncertainty of fertility risk stratification and the importance of early fertility counseling and individualized fertility preservation strategies. To illustrate the potential clinical impact, we present the case of a 34-year-old nulliparous woman who developed premature ovarian insufficiency two years after neoadjuvant chemoimmunotherapy including atezolizumab, despite ovarian suppression. In conclusion, while ICIs have transformed the therapeutic landscape of TNBC, their potential long-term impact on reproductive and endocrine health represents a clinically significant concern. A precautionary, multidisciplinary oncofertility approach and prospective clinical registries are essential to define the true incidence and mechanisms of ICI-associated reproductive toxicity. Full article

Background: Rice is one of the world’s main staple crops , and improving its productivity and resilience is important to achieving food security under varying climatic conditions. Objectives: This systematic review synthesizes the existing evidence on the application, technical limitations, and potential of the development of genome editing technologies (CRISPR-Cas) in rice (Oryza sativa L.), as well as presents a novel approach called the CRISPR Trait Prioritization and Readiness Framework (CTPRF). Methods: Peer-reviewed articles that reported applications of genome editing based on the CRISPR-Cas system in the genome of rice for trait improvement or functional genomics were identified through searches fromPubMed, Scopus, Web of Science, and Google Scholar with studies published between 2012 and 2025. Studies were screened on predefined inclusion criteria related to experimental validation, reporting of editing efficiency, and clear phenotypic results. Data on CRISPR systems, target genes, methods of delivery, traits modified, and phenotypic results were extracted and synthesized by comparative analysis. Results: A wide variety of different CRISPR systems have been used in rice, and our results indicate that NHEJ-mediated knockouts are effective in average genotypes with editing efficiencies in the range of 70–90%, but HDR and prime editing are still under 10%. The CTPRF is being introduced as a strategic decision support tool to evaluate traits from four dimensions: technical feasibility, phenotypic predictability, impact potential, and regulatory pathway. We use this framework for case studies in pioneering countries (USA, Japan, China) and show how it can be useful for guiding research investment and policy. Conclusions: CRISPR-Cas technologies have transformed rice breeding, but their introduction requires overcoming genotype-dependent barriers to transformation and negotiating patchwork regulatory environments. The CTPRF offers a roadmap for the acceleration of the development of climate-resilient and nutritious rice varieties for the action plan. Full article

To enhance the autonomous navigation capability of intelligent agents in complex environments, this paper presents a visual localization algorithm for dynamic scenes that leverages multi-source information fusion. The proposed approach is built upon an odometry framework integrating LiDAR, camera, and IMU data, and incorporates the YOLOv8 model to extract semantic information from images, which is then fused with laser point cloud data. We design a dynamic point removal method based on multi-modal association, which links 2D image masks to 3D point cloud regions, applies Euclidean clustering to differentiate static and dynamic points, and subsequently employs PnP-RANSAC to eliminate any remaining undetected dynamic points. This process yields a robust localization algorithm for dynamic environments. Experimental results on datasets featuring dynamic objects and a custom-built hardware platform demonstrate that the proposed dynamic point removal method significantly improves both the robustness and accuracy of the visual localization system. These findings confirm the feasibility and effectiveness of our system, showcasing its capabilities in precise positioning and autonomous navigation in complex environments. Full article

Legged robot locomotion remains a critical challenge in robotics, demanding control strategies that are not only dynamically stable and robust but also capable of adapting to complex and changing environments. deep reinforcement learning (DRL) has recently emerged as a powerful approach to automatically generate motion control policies by learning from interactions with simulated or real environments. This study provides a systematic overview of DRL applications in legged robot control, emphasizing experimental platforms, measurement techniques, and benchmarking practices. Following PRISMA guidelines, 27 peer-reviewed studies published between 2018 and 2025 were analyzed, covering model-free, model-based, hierarchical, and hybrid DRL frameworks. Our findings reveal that reward shaping, policy representation, and training stability significantly influence control performance, while domain randomization and dynamic adaptation methods are essential for bridging the simulation-to-real-world gap. In addition, this review highlights instrumentation approaches for evaluating algorithm effectiveness, offering insights into sample efficiency, energy management, and safe deployment. The results aim to guide the development of reproducible and experimentally validated DRL-based control systems for legged robots. Full article

Heterojunctions combining graphene with transition metal dichalcogenides (TMDCs) have garnered considerable interest in phototransistor research. Molybdenum disulfide (MoS₂) can be well combined with graphene owing to its excellent and special bandgap characteristics. In this study, a photoelectric transistor is designed and fabricated based on a graphene–molybdenum disulfide (MoS₂) van der Waals heterojunction. Its novelty lies in constructing a vertical heterojunction architecture with a well-defined structure, clear interface, and easy gate modulation. It fully utilizes the high mobility of graphene and the appropriate bandgap of MoS₂ to achieve efficient light absorption and carrier transport. The device exhibits a good photoelectric response and stability at room temperature, with key performance indicators including the following: a responsivity of 0.5023 mA/W, and a dark current of approximately 10⁻¹¹ A at a gate voltage of 0 V and approaching 10⁻¹⁰ A at 30 V; when the light intensity is 1000 mW/cm², the photocurrent reaches the 10⁻⁸ A level, demonstrating the synergistic modulation capability of gate voltage and light intensity. Although its responsivity is lower than some high-performance heterojunction devices, this device has advantages such as a simple structure, controllable preparation, stable room-temperature operation, and the potential for a broad-spectrum response, showing good application prospects in flexible electronics and integrated optoelectronic systems. This study provides an experimental basis and technical path for the development of two-dimensional material heterojunctions in programmable, multifunctional optoelectronic devices. Full article