Search Results (1,003)

The immersive experience of holographic displays is fundamentally limited by their space–bandwidth product (SBP), which imposes an inherent trade-off between the field of view (FOV) and eyebox size. This paper proposes a method to extend the SBP by employing cascaded modulation with a dynamic spatial light modulator (SLM) and a passive high-resolution binary random phase mask (BRPM). We find that the key to unlocking this extension of SBP lies in a sophisticated algorithmic optimization, grounded in a physically accurate model of the system. We identify and correct the Nyquist undersampling problem caused by high-frequency scattering in standard diffraction models. Based on this physically accurate model, we employ a gradient descent optimization framework to achieve efficient, end-to-end solving for complex light fields. Simulation and experimental results demonstrate that our method achieves an approximately 16-fold SBP extension (4-fold FOV) while delivering significantly superior reconstructed image quality compared to the traditional Gerchberg–Saxton (GS) algorithm. Furthermore, this study quantitatively reveals the system’s extreme sensitivity to sub-pixel level alignment accuracy, providing critical guidance for the engineering and implementation of our proposed method. Full article

Contemporary visual assistive devices often lack immersive user experience due to passive control systems. This study introduces a neuronally controlled visual assistive device (NCVAD) that aims to assist visually impaired users in performing reach tasks with active, intuitive control. The developed NCVAD integrates computer vision, electroencephalogram (EEG) signal processing, and robotic manipulation to facilitate object detection, selection, and assistive guidance. The monocular vision-based subsystem implements the YOLOv8n algorithm to detect objects of daily use. Then, audio prompting conveys the detected objects’ information to the user, who selects their targeted object using a voluntary trigger decoded through real-time EEG classification. The target’s physical coordinates are extracted using ArUco markers, and a gradient descent-based path optimization algorithm (POA) guides a 3-DoF robotic arm to reach the target. The classification algorithm achieves over 85% precision and recall in decoding EEG data, even with coexisting physiological artifacts. Similarly, the POA achieves approximately 650 ms of actuation time with a 0.001 learning rate and 0.1 cm² error threshold settings. In conclusion, the study also validates the preliminary analysis results on a working physical model and benchmarks the robotic arm’s performance against human users, establishing the proof-of-concept for future assistive technologies integrating EEG and computer vision paradigms. Full article

Scanned data often contain substantial outliers due to environmental interference, which drastically decreases the performance of traditional registration algorithms. To address this issue, this article proposes an improved principal component analysis (PCA) and Jacobian-enhanced whale optimization collaborative method for point cloud registration. First, an improved PCA point cloud initial registration algorithm is proposed by introducing the normal vector local information to set the screening conditions. This algorithm can streamline the original set of 48 candidate rotation matrices down to 4, achieving rapid point cloud registration at the data level between the scanned and model point clouds. Second, a Jacobian whale optimization algorithm for fine registration (JWOA-FR) is proposed by incorporating local gradient information. The algorithm employs gradient descent on optimal whale individuals to dynamically guide global search updates, thereby enhancing both registration accuracy and efficiency. Finally, a threshold is set to remove the outliers contained in the workpieces based on the information of the matched point pairs. The iterative closest point (ICP) algorithm is further used to improve registration accuracy for data without outliers. The experimental results showed that registration errors of large workpieces 1, 2, and 3 were 2.0755 mm, 2.3955 mm, and 2.5823 mm, respectively, after outlier removal, which indicates that the proposed method is applicable to data with outliers, and the registration accuracy meets the requirements. Full article

Generative video diffusion models (GVDs) generate high-fidelity, text-conditioned videos but risk producing unsafe or copyrighted content due to training on large, uncurated datasets. Concept erasure techniques aim to remove such harmful concepts from pre-trained models while preserving overall generative performance. However, existing methods mainly target single-concept erasure and thus cannot satisfy the demand for simultaneously eliminating multi-concept in real-world scenarios. On the one hand, naively applying single-concept erasure sequentially to multi-concept often yields suboptimal results due to conflicts among target concepts; on the other hand, methods that alter concept mappings exhibit very poor adaptability and fail to accommodate the dynamic concept changes. To address these, we propose ConceptVoid, a scalable multi-concept erasure framework formulated as a constrained multi-objective optimization problem. For each target concept, an erasure loss is defined as the discrepancy between noise predictions conditioned and unconditioned on the concept. Non-target generation capabilities are preserved via output-distribution alignment regularization. We apply the multiple gradient descent algorithm (MGDA) to obtain Pareto-optimal solutions, aiming to minimize conflicts among different concept erasure objectives. In addition, we improve MGDA by introducing an importance-weighting mechanism, which adjusts the weights of gradients corresponding to each erasure objective, enabling flexible control over the priority and intensity of erasing different concepts, thereby enhancing the scalability of ConceptVoid. Extensive experiments demonstrate the effectiveness of ConceptVoid, validating our key contributions: (1) a scalable framework for multi-concept erasure in GVDs; (2) the integration of per-concept erasure with distribution alignment to retain non-target quality; and (3) an enhanced MGDA for conflict-aware, controllable erasure. Full article

The efficient utilization of structural information in High-Range Resolution Profiles (HRRPs) is of great significance for improving recognition performance. This paper proposes a size estimation method based on L1-norm variable fractional-order gradient descent, which achieves size inversion in complex electromagnetic environments by establishing an HRRP projection model of ballistic targets. Specifically: First, through rigorous geometrical optics analysis, an analytical relationship model between the target’s projected size and actual size is established. Second, an error function under the L1-norm is constructed, and an adaptive order-adjusting fractional-order gradient descent method is employed for optimization, effectively overcoming the sensitivity to outliers inherent in traditional L2-norm methods. Finally, by introducing a dynamic order-switching mechanism, computational efficiency is improved while ensuring convergence accuracy. Experimental results show that at a measurement error of 0.4 m, the proposed method maintains excellent estimation performance with sensitivity to outliers reduced, and the actual size inversion error remains stable below 3.7%. Full article

Machine learning (ML) has greatly improved intrusion detection in enterprise networks. However, ML models remain vulnerable to adversarial attacks, where small input changes cause misclassification. This study evaluates the robustness of a Random Forest (RF), a standard neural network (NN), and a Transformer-based Network Intrusion Detection System (NIDS). It also introduces ADV_NN, an adversarially trained neural network designed to improve resilience. Model performance is tested using the UNSW-NB15 dataset under both clean and adversarial conditions. The attack types include Projected Gradient Descent (PGD), Fast Gradient Sign Method (FGSM), and Black-Box transfer attacks. The proposed ADV_NN achieves 86.04% accuracy on clean data. It maintains over 80% accuracy under PGD and FGSM attacks, and exceeds 85% under Black-Box attacks at $ϵ=0.15$. In contrast, the RF, NN, and Transformer-based models suffer significant degradation under adversarial perturbations. These results highlight the need to incorporate adversarial defenses into ML-based NIDS for secure deployment in real-world environments. Full article

Background/Objectives: Accurate prediction of pediatric appendicitis diagnosis, management, and severity is critical for clinical decision-making. We aimed to evaluate the predictive performance of a wide range of machine learning models, combined with various feature selection techniques, on a pediatric appendicitis dataset. A particular focus was placed on the role of ultrasound (US) image-descriptive features in model performance and explainability. Methods: We conducted a retrospective cohort study on a dataset of 781 pediatric patients aged 0–18 presenting to Children’s Hospital St. Hedwig in Regensburg, Germany, between January 2016 and February 2023. We developed and validated predictive models; machine learning algorithms included the random forest, logistic regression, stochastic gradient descent, and the light gradient boosting machine (LGBM). These were paired exhaustively with feature selection methods spanning filter-based (association and prediction), embedded (LGBM and linear), and a novel redundancy-aware step-up wrapper approach. We employed a machine learning benchmarking study design where AI models were trained to predict diagnosis, management, and severity outcomes, both with and without US image-descriptive features, and evaluated on held-out testing samples. Model performance was assessed using overall accuracy and area under the receiver operating characteristic curve (AUROC). A deep learner optimized for tabular data, GANDALF, was also evaluated in these applications. Results: US features significantly improved diagnostic accuracy, supporting their use in reducing model bias. However, they were not essential for maximizing accuracy in predicting management or severity. In summary, our best-performing models were, for diagnosis, the random forest with embedded LGBM feature selection (98.1% accuracy, AUROC: 0.993), for management, the random forest without feature selection (93.9% accuracy, AUROC: 0.980), and for severity, the LGBM with filter-based association feature selection (90.1% accuracy, AUROC: 0.931). Conclusions: Our results demonstrate that high-performing, interpretable machine learning models can predict key clinical outcomes in pediatric appendicitis. US image features improve diagnostic accuracy but are not critical for predicting management or severity. Full article

Unmanned aerial vehicles (UAVs) play an ever-increasing role in disaster response and remote sensing. However, the deep learning models they rely on remain highly vulnerable to adversarial attacks. This paper presents an evaluation and defense framework aimed at enhancing adversarial robustness in aerial disaster image classification using the AIDERV2 dataset. Our methodology is structured into the following four phases: (I) baseline training with clean data using ResNet-50, (II) vulnerability assessment under Projected Gradient Descent (PGD) attacks, (III) adversarial training with PGD to improve model resilience, and (IV) comprehensive post-defense evaluation under identical attack scenarios. The baseline model achieves 93.25% accuracy on clean data but drops to as low as 21.00% under strong adversarial perturbations. In contrast, the adversarially trained model maintains over 75.00% accuracy across all PGD configurations, reducing the attack success rate by more than 60%. We introduce metrics, such as Clean Accuracy, Adversarial Accuracy, Accuracy Drop, and Attack Success Rate, to evaluate defense performance. Our results show the practical importance of adversarial training for safety-critical UAV applications and provide a reference point for future research. This work contributes to making deep learning systems on aerial platforms more secure, robust, and reliable in mission-critical environments. Full article

Network trace is a comprehensive record of data packets traversing a computer network, serving as a critical resource for analyzing network behavior. However, in practice, the limited availability of high-quality network traces, coupled with the presence of sensitive information such as IP addresses and MAC addresses, poses significant challenges to advancing network trace analysis. To address these issues, this paper focuses on network trace synthesis in two practical scenarios: (1) data expansion, where users create synthetic traces internally to diversify and enhance existing network trace utility; (2) data release, where synthesized network traces are shared externally. Inspired by the powerful generative capabilities of latent diffusion models (LDMs), this paper introduces NetSynDM, which leverages LDM to address the challenges of network trace synthesis in data expansion scenarios. To address the challenges in the data release scenario, we integrate differential privacy (DP) mechanisms into NetSynDM, introducing DPNetSynDM, which leverages DP Stochastic Gradient Descent (DP-SGD) to update NetSynDM, incorporating privacy-preserving noise throughout the training process. Experiments on five widely used network trace datasets show that our methods outperform prior works. NetSynDM achieves an average 166.1% better performance in fidelity compared to baselines. DPNetSynDM strikes an improved balance between privacy and fidelity, surpassing previous state-of-the-art network trace synthesis method fidelity scores of 18.4% on UGR16 while reducing privacy risk scores by approximately 9.79%. Full article

Large language models (LLMs) generally perform well in common tasks, yet are often susceptible to errors in sophisticated natural language processing (NLP) on classification applications. Prompt engineering has emerged as a strategy to enhance their performance. Despite the effort required for manual prompt optimization, recent advancements highlight the need for automation to reduce human involvement. We introduced the PO2G (prompt optimization with two gradients) framework to improve the efficiency of optimizing prompts for classification tasks. PO2G demonstrates improvement in efficiency, reaching almost 89% accuracy after just three iterations, whereas ProTeGi requires six iterations to achieve a comparable level. We evaluated PO2G and ProTeGi on a benchmark of nine NLP tasks, three tasks from the original ProTeGi study, and six non-domain-specific tasks. We also evaluated both frameworks on seven legal-domain classification tasks. These results provide broader insights into the efficiency and effectiveness of prompt optimization frameworks for classification across diverse NLP scenarios. Full article

We study the generalization properties of stochastic optimization methods under adaptive data sampling schemes, focusing on the setting of pairwise learning, which is central to tasks like ranking, metric learning, and AUC maximization. Unlike pointwise learning, pairwise methods must address statistical dependencies between input pairs—a challenge that existing analyses do not adequately handle when sampling is adaptive. In this work, we extend a general framework that integrates two algorithm-dependent approaches—algorithmic stability and PAC–Bayes analysis for this purpose. Specifically, we examine (1) Pairwise Stochastic Gradient Descent (Pairwise SGD), widely used across machine learning applications, and (2) Pairwise Stochastic Gradient Descent Ascent (Pairwise SGDA), common in adversarial training. Our analysis avoids artificial randomization and leverages the inherent stochasticity of gradient updates instead. Our results yield generalization guarantees of order $n^{-1/2}$ under non-uniform adaptive sampling strategies, covering both smooth and non-smooth convex settings. We believe these findings address a significant gap in the theory of pairwise learning with adaptive sampling. Full article

Detecting steel defects is a vital process in industrial production, but traditional methods suffer from poor feature extraction and low detection accuracy. To address these issues, this research introduces an improved model, EB-YOLOv8, based on YOLOv8. First, the multi-scale attention mechanism EMA is integrated into the backbone and neck sections to reduce noise during gradient descent and enhance model stability by encoding global information and weighting model parameters. Second, the weighted fusion splicing module, Concat_BiFPN, is used in the neck network to facilitate multi-scale feature detection and fusion. This improves detection precision. The results show that the EB-YOLOv8 model increases detection accuracy on the NEU-DET dataset by 3.1%, reaching 80.2%, compared to YOLOv8. Additionally, the average precision on the Severstal steel defect dataset improves from 65.4% to 66.1%. Overall, the proposed model demonstrates superior recognition performance. Full article

Due to the complex spatial structure, dense tree distribution, and narrow passages in orchard environments, traditional path planning algorithms often struggle with large path deviations, frequent turning, and reduced navigational safety. In order to overcome these challenges, this paper proposes a hybrid path planning algorithm based on improved D* Lite for narrow forest orchard environments. The proposed approach enhances path feasibility and improves the robustness of the navigation system. The algorithm begins by constructing a 2D grid map reflecting the orchard layout and inflates the tree regions to create safety buffers for reliable path planning. For global path planning, an enhanced D* Lite algorithm is used with a cost function that jointly considers centerline proximity, turning angle smoothness, and directional consistency. This guides the path to remain close to the orchard row centerline, improving structural adaptability and path rationality. Narrow passages along the initial path are detected, and local replanning is performed using a Hybrid A* algorithm that accounts for the kinematic constraints of a differential tracked robot. This generates curvature-continuous and directionally stable segments that replace the original narrow-path portions. Finally, a gradient descent method is applied to smooth the overall path, improving trajectory continuity and execution stability. Field experiments in representative orchard environments demonstrate that the proposed hybrid algorithm significantly outperforms traditional D* Lite and KD* Lite-B methods in terms of path accuracy and navigational safety. The average deviation from the centerline is only 0.06 m, representing reductions of 75.55% and 38.27% compared to traditional D* Lite and KD* Lite-B, respectively, thereby enabling high-precision centerline tracking. Moreover, the number of hazardous nodes, defined as path points near obstacles, was reduced to five, marking decreases of 92.86% and 68.75%, respectively, and substantially enhancing navigation safety. These results confirm the method’s strong applicability in complex, constrained orchard environments and its potential as a foundation for efficient, safe, and fully autonomous agricultural robot operation. Full article

To address the issues of DC-side voltage fluctuation and three-phase current distortion in rectifier systems under pulsed load conditions, this paper proposes a control strategy that integrates Model Predictive Control (MPC) with a Luenberger observer for the Power Pulsation Buffer (PPB). The observer parameters are adaptively tuned using a gradient descent method. First, the pulsed current generated by the load is decomposed into dynamic and average components, and a mathematical model of the PPB is established. Considering the negative impact of DC voltage ripple and lumped disturbances such as parasitic parameters on model accuracy, a Luenberger observer is designed to estimate these disturbances. To overcome the dependence of traditional Luenberger observers on empirically tuned gains, an adaptive gradient descent algorithm based on gradient direction consistency is introduced for online gain adjustment. Simulation and experimental results demonstrate that the proposed control strategy—combining the Luenberger observer with gradient descent and MPC—effectively reduces current tracking overshoot and improves tracking accuracy. Furthermore, it enables sustained decoupling of the PPB from the system, significantly mitigating DC-side voltage ripple and three-phase current distortion under pulsed load conditions, thereby validating the effectiveness of the proposed approach. Full article

Effective earthquake early warning (EEW) is essential for disaster prevention in the built environment, enabling a rapid structural response, system shutdown, and occupant evacuation to mitigate damage and casualties. However, most current EEW systems lack rigorous reliability analyses of their predictive outcomes, limiting their effectiveness in real-world scenarios—especially for on-site warnings, where data are limited and time is critical. To address these challenges, we propose a Bayesian neural network (BNN) framework based on Stein variational gradient descent (SVGD). By performing Bayesian inference, we estimate the posterior distribution of the parameters, thus outputting a reliability analysis of the prediction results. In addition, we incorporate a continual learning mechanism based on elastic weight consolidation, allowing the system to adapt quickly without full retraining. Our experiments demonstrate that our SVGD-BNN model significantly outperforms traditional peak displacement (Pd)-based approaches. In a 3 s time window, the Pearson correlation coefficient R increases by 9.2% and the residual standard deviation $SD$ decreases by 24.4% compared to a variational inference (VI)-based BNN. Furthermore, the prediction variance generated by the model can effectively reflect the uncertainty of the prediction results. The continual learning strategy reduces the training time by 133–194 s, enhancing the system’s responsiveness. These features make the proposed framework a promising tool for real-time, reliable, and adaptive EEW—supporting disaster-resilient building design and operation. Full article