Search Results (109)

Research on deep neural network (DNN)-based multi-dimensional data visualization has thoroughly explored cross-modal hash retrieval (CMHR) systems, yet their vulnerability to malicious adversarial examples remains evident. Recent work improves the robustness of CMHR networks by augmenting training datasets with adversarial examples. Prior approaches typically formulate the generation of cross-modal adversarial examples as an optimization problem solved through iterative methods. Although effective, such techniques often suffer from slow generation speed, limiting research efficiency. To address this, we propose a generative-based method that enables rapid synthesis of adversarial examples via a carefully designed adversarial generator network. Specifically, we introduce Cross-Gen, a parallel cross-modal framework that constructs semantic triplet data by interacting with the target model through query-based feedback. The generator is optimized using a tailored objective comprising adversarial loss, reconstruction loss, and quantization loss. The experimental results show that Cross-Gen generates adversarial examples significantly faster than iterative methods while achieving competitive attack performance. Full article

Teaching-based robotic systems offer an accessible alternative to complex inverse kinematics programming for food processing automation. Traditional model-based approaches require precise system identification and analytical solutions that are challenging for custom-built robots with manufacturing tolerances and mechanical uncertainties. This study developed a custom six-degree-of-freedom robotic arm using modular brushless motors controlled via Controller Area Network communication and Robot Operating System 2, a teaching mode where users manually demonstrate trajectories that are recorded at 100 Hz. Forty-five demonstration trajectories were collected across three geometric patterns (rectangle, triangle, circle) and augmented to 270 samples. A bidirectional Long Short-Term Memory network with attention mechanism was trained to classify patterns, achieving 83.33% test accuracy and outperforming baseline deep learning models (1D-CNN: 77.78%, TCN: 66.67%, GRU: 44.44%), while being marginally exceeded by Random Forest (86.11%). Rectangle patterns showed strongest recognition (78.57% F1-score), while circle patterns achieved highest performance (91.67% F1-score). However, severe overfitting was observed, with validation accuracy peaking at 85.19% at epoch 24 before degradation, indicating insufficient dataset size despite five-fold augmentation. The results demonstrate proof-of-concept feasibility for pattern recognition from limited teaching demonstrations, providing a pathway for robotic food processing without extensive programming expertise, though larger datasets and robust feedback control strategies are required for production deployment. Full article

English argumentative writing is a cornerstone of academic and professional communication, yet it remains a significant challenge for second-language (L2) learners. While Large Language Models (LLMs) show promise as components in automated feedback systems, their responses are often generic and lack the structural insight necessary for meaningful improvement. Existing Automated Essay Scoring (AES) systems, conversely, typically provide holistic scores without the kind of actionable, fine-grained advice that can guide concrete revisions. To bridge this systemic gap, we introduce ARGUS (Argument Understanding and Structured-feedback), a novel neuro-symbolic system that synergizes the semantic understanding of LLMs with the structured reasoning of Graph Neural Networks (GNNs). The ARGUS system architecture comprises three integrated modules: (1) an LLM-based parser transforms an essay into a structured argument graph; (2) a Relational Graph Convolutional Network (R-GCN) analyzes this symbolic structure to identify specific logical and structural flaws; and (3) this flaw analysis directly guides a conditional LLM to generate feedback that is not only contextually relevant but also pinpoints precise weaknesses in the student’s reasoning. We evaluate ARGUS on the Argument Annotated Essays corpus and on an additional set of 150 L2 persuasive essays collected from the same population to augment training of both the parser and the structural flaw detector. Our argument parsing module achieves a component identification F1-score of 90.4% and a relation identification F1-score of 86.1%. The R-GCN-based structural flaw detector attains a macro-averaged F1-score of 0.83 across the seven flaw categories, indicating that the enriched training data substantially improves its generalization. Most importantly, in a human evaluation study, feedback generated by the ARGUS system was rated as consistently and significantly more specific, accurate, actionable, and helpful than that from strong baselines, including a fine-tuned LLM and a zero-shot GPT-4. Our work demonstrates a robust systems engineering approach, grounding LLM-based feedback in GNN-driven structural analysis to create an intelligent teaching system that provides targeted, pedagogically valuable guidance for L2 student writers engaging with persuasive essays. Full article

Background: Artificial intelligence (AI) and machine learning (ML) have rapidly expanded across the continuum of cardiopulmonary resuscitation (CPR), with growing evidence of their contribution to improving early recognition, intervention quality, and post-cardiac arrest outcomes. This narrative review synthesizes the current advancements and challenges in AI/ML-enhanced resuscitation science. Methods: A targeted literature search was conducted in Web of Science for the period 2018–2025 using the keywords “artificial intelligence” and “cardiopulmonary resuscitation”. The search identified studies addressing AI/ML applications across the resuscitation pathway, which were reviewed and categorized according to the American Heart Association’s Chain of Survival—prevention and preparedness, activation of the emergency response system, high-quality CPR including early defibrillation, advanced resuscitation interventions, post-cardiac arrest care, and recovery. Results: The literature demonstrates substantial promise for AI/ML in several domains: (1) early recognition and timely activation of emergency medical services through real-time detection algorithms; (2) optimization of high-quality CPR, including feedback systems, automated assessment of chest compressions, and prediction of defibrillation success; (3) support for advanced resuscitation interventions, such as rhythm classification, prognostication, and intra-arrest decision support; (4) post-cardiac arrest care, including outcome prediction and neuroprognostication; and (5) integrative and cross-domain approaches that link multiple phases of resuscitation into end-to-end AI-supported systems. Emerging work also highlights the role of AI in education and training, with applications in simulation, assessment, and skill reinforcement. Conclusions: AI/ML technologies hold significant potential to augment clinical performance across all links of the Chain of Survival. Their effective implementation requires attention to ethical considerations, data representativeness, and real-world validation. Future research should prioritize multicenter datasets, transparency, bias mitigation, and clinically embedded evaluation frameworks to ensure that AI/ML systems support safe, equitable, and high-impact resuscitation care. Full article

Background: Knee osteoarthritis (OA) is a prevalent musculoskeletal condition associated with pain, functional limitation, and reduced quality of life. Telerehabilitation has emerged as a scalable intervention, yet many platforms lack neuromuscular feedback or objective-monitoring capabilities. The KneE-PAD system uniquely integrates electromyographic and inertial sensing to provide personalized feedback and remote performance tracking. Objective: To evaluate the clinical effectiveness of a sensor-augmented telerehabilitation system (KneE-PAD) compared to conventional face-to-face physiotherapy in older adults with mild-to-moderate knee OA. Methods: In this single-blind randomized controlled trial, 42 older adults (mean age 68.4 ± 5.7 years) were randomly assigned to either KneE-PAD telerehabilitation or conventional physiotherapy for eight weeks. KneE-PAD sessions incorporated real-time electromyographic and motion feedback, while physiotherapists remotely supervised training. Assessments were performed at baseline, post-intervention, and 12-week follow-up. Primary outcomes included quadriceps strength, neuromuscular activation, and WOMAC scores. Secondary outcomes covered functional mobility, psychological distress, self-efficacy, and fear of movement. Results: The telerehabilitation group demonstrated notable improvements in neuromuscular activation, quadriceps strength, and functional capacity, all exceeding clinically meaningful thresholds. Functional mobility and pain outcomes showed substantial gains compared with the control group, while psychological indicators (self-efficacy and depressive symptoms) exhibited modest but positive trends. Between-group comparisons consistently favored KneE-PAD, with effects maintained at the 12-week follow-up, confirming both clinical and functional robustness. Conclusions: Sensor-augmented telerehabilitation using the KneE-PAD platform appears to be a feasible and potentially effective alternative to conventional physiotherapy for knee OA. By combining real-time feedback, motor learning reinforcement, and remote monitoring, the system may enhance neuromuscular and functional recovery. These findings should be confirmed in larger and longer-term trials. Trial Registration: ClinicalTrials.gov: NCT06416332. Full article

Spinal cord injury (SCI) impairs motor function and requires rigorous rehabilitative therapy, motivating the development of approaches that are engaging and customizable. Virtual reality (VR) motor training with augmented sensory feedback (ASF) offers a promising pathway to enhance functional outcomes, yet it remains unclear how ASF modalities affect performance and underlying psychophysiological states in persons with SCI. Five participants with chronic incomplete cervical-level SCI controlled a virtual robotic arm with semi-isometric upper-body contractions while undergoing ASF training with either visual feedback (VF) or combined visual plus haptic feedback (VHF). Motor performance (pathlength, completion time), psychophysiological measures (EEG, EMG, EDA, HR), and perceptual ratings (agency, motivation, utility) were assessed before and after ASF training. VF significantly reduced pathlength (−12.5%, p = 0.0011) and lowered EMG amplitude (−32.5%, p = 0.0063), suggesting the potential for improved motor performance and neuromuscular efficiency. VHF did not significantly improve performance, but trended toward higher cortical engagement. EEG analyses showed VF significantly decreased alpha and beta activity after training, whereas VHF trended toward mild increases. Regression revealed improved performance was significantly (p < 0.05) associated with changes in alpha power, EMG, EDA, and self-reported motivation. ASF type may differentially shape performance and psychophysiological responses in SCI participants. These preliminary findings suggest VR-based ASF as a potent multidimensional tool for personalizing rehabilitation. Full article

Recycling industrial solid waste phosphogypsum into phosphogypsum concrete (PGC) is a crucial pathway for achieving high-value solid waste utilization. However, the scarcity of experimental samples for PGC has led to inaccurate predictions of compressive strength by traditional models, severely hindering its application. This study proposes a dynamic weighted transfer learning-based method for predicting the strength of PGC, addressing the characterization bottleneck under small-sample conditions by transferring knowledge from the strength patterns of conventional concrete. First, feature differences between conventional concrete and PGC are eliminated through component proportion normalization and feature alignment. Then, a data augmentation technique based on Bootstrap Resampling is developed to generate enhanced samples that comply with mix proportion constraints, effectively expanding the training samples. Finally, an error feedback-driven dynamic weight calculation and weighted loss optimization framework for transfer learning is designed, prioritizing the learning of samples in the prediction blind spots of the target domain. This enables the adaptive acquisition of PGC-specific knowledge while inheriting the general knowledge of conventional concrete. Experimental results show that the transfer learning model achieves a prediction accuracy of R² = 0.95 on the target domain test samples, a 15.9% improvement over traditional methods, while maintaining robust performance (R² = 0.97) on an external validation samples. Feature importance analysis and Shapley Additive Explanations (SHAP) analysis reveal the nonlinear coupling effects of PGC-specific parameters on strength. This study establishes a scientific approach for accurate strength prediction of PGC under small-sample conditions. Full article

Background: Emotionally salient music may enhance attention-focused rehabilitation, yet concurrent music plus virtual-reality programs in chronic stroke are largely untested. We assessed whether personalized emotional music stimulation (EMS) layered onto a standardized virtual reality rehabilitation system (VRRS) augments cognitive, affective, physiological, and functional outcomes. Methods: In a quasi-randomized outpatient trial, 20 adults ≥ 6 months post-ischemic stroke were allocated by order of recruitment to VRRS alone (control, n = 10) or VRRS+EMS (experimental, n = 10). Both groups performed 45 min of active VRRS cognitive training (3×/week, 8 weeks), while the EMS group received approximately 60 min sessions including setup and feedback phases. Primary outcomes were cognition and global function; secondary outcomes were intrinsic motivation, depression, anxiety, and heart rate. Non-parametric tests with effect sizes and Δ-scores were used. Results: The experimental group improved across all domains: cognition (median +4.5 points), motivation (median +54 points), depression (median −3.5 points), anxiety (median −4.0 points), heart rate (median −6.35 beats per minute), and disability (median one-grade improvement), each with large effects. The control group showed smaller gains in cognition and motivation and a modest heart-rate reduction, without significant changes in mood or disability. At post-treatment, the music group outperformed controls on cognition, motivation, and disability. Change-score analyses favored the music group for every endpoint. Larger heart-rate reductions correlated with greater improvements in depression (ρ = 0.73, p < 0.001) and anxiety (ρ = 0.58, p = 0.007). Conclusions: Adding personalized emotional music to virtual-reality attention training produced coherent, clinically relevant gains in cognition, mood, motivation, autonomic regulation, and independence compared with virtual reality alone. Full article

Background: Electroencephalography (EEG) has been widely used in Attention Deficit Hyperactivity Disorder (ADHD) recognition, but existing methods still suffer from limitations in dynamic modeling, small-sample adaptability, and training stability. This study proposes a Multi-Source Fusion and Feedback Network (MSRLNet) to enhance EEG-based ADHD recognition. Methods: MSRLNet comprises three modules: (1) Multi-Source Feature Fusion (MSFF), combining microstate and statistical features to improve interpretability; (2) a CNN-GRU Parallel Module (CGPM) for multi-scale temporal modeling; and (3) Performance Feedback–driven Parameter Optimization (PFPO) to enhance training stability. Feature-level data augmentation is introduced to alleviate overfitting in small-sample scenarios. Results: On a public dataset, MSRLNet achieved an accuracy of 98.90%, an F1-score of 98.98%, and a kappa of 0.979, all exceeding comparative approaches. Conclusions: MSRLNet shows high accuracy and robustness in ADHD EEG feature recognition, verifying its potential application value in clinical auxiliary diagnosis. Full article

Vascular surgery has evolved from durable yet invasive open reconstructions to less traumatic endovascular techniques. While endovascular repair reduces perioperative morbidity, it introduces durability challenges and the need for lifelong surveillance. Laparoscopic surgery bridged some gaps but was hindered by steep learning curves and technical limitations. Robotic-assisted surgery represents a “third revolution”, combining the durability of open repair with the recovery and ergonomic benefits of minimally invasive approaches through enhanced 3D visualization, wristed instrumentation, and tremor filtration. This review synthesizes current evidence on robotic applications in vascular surgery, including aortic, visceral, venous, and endovascular interventions. Feasibility of robotic vascular surgery has been demonstrated in over 1500 patients across aortic, visceral, venous, and decompression procedures. Reported outcomes include pooled conversion rates of ~5%, 30-day mortality of 1–3%, and long-term patency rates exceeding 90% in aortoiliac occlusive disease. Similarly favorable outcomes have been observed in AAA repair, visceral artery aneurysm repair, IVC reconstructions, renal vein transpositions, and minimally invasive decompression procedures such as median arcuate ligament and thoracic outlet syndromes. Endovascular robotics enhances catheter navigation precision and reduces operator radiation exposure by 85–95%, with multiple series demonstrating consistent benefit compared to manual techniques. Despite these advantages, adoption is limited by high costs, lack of dedicated vascular instruments, absent haptic feedback on most platforms, and the need for standardized training. Most available evidence is observational and from high-volume centers, highlighting the need for multicenter randomized trials. Future directions include AI-enabled planning and augmented-reality navigation, which are the most feasible near-term technologies since they rely largely on software integration with existing systems. Other advances such as microsurgical robotics, soft-robotic platforms, and telesurgery remain longer-term developments requiring new hardware and regulatory pathways. Overcoming barriers through collaborative innovation, structured training, and robust evidence generation is essential for robotics to become a new standard in vascular care. Full article

The integration of Virtual Worlds (VW) and Intelligent Tutoring Systems (ITS) represents a transformative advancement in education, combining immersive, interactive learning with AI-driven personalization. This study explores the synergies between these technologies, analyzing their benefits, challenges, and applications in domains such as medical training, STEM education, and language learning. Findings highlight their shared characteristics of adaptability, real-time feedback, and collaborative learning. However, challenges such as computational demands, pedagogical complexity, and ethical concerns must be addressed. Future research should focus on hybrid models leveraging blockchain, IoT, and augmented reality to enhance adaptive and scalable learning experiences. Full article

Edge intelligence plays an increasingly vital role in ensuring the reliability of distributed microservice-based applications, which are widely used in domains such as e-commerce, industrial IoT, and cloud-edge collaborative platforms. However, anomaly detection in these systems encounters a critical challenge: labeled anomaly data are scarce. This scarcity leads to severe class asymmetry and compromised detection performance, particularly under the resource constraints of edge environments. Recent approaches based on Graph Neural Networks (GNNs)—often integrated with DeepSVDD and regularization techniques—have shown potential, but they rarely address this asymmetry in an adaptive, scenario-specific way. This work proposes Heterogeneous Graph Adaptive Augmentation (HGAA), a framework tailored for edge intelligence scenarios. HGAA dynamically optimizes graph data augmentation by leveraging feedback from online anomaly detection. To enhance detection accuracy while adhering to resource constraints, the framework incorporates a selective bias toward underrepresented anomaly types. It uses knowledge distillation to model dataset-dependent distributions and adaptively adjusts augmentation probabilities, thus avoiding excessive computational overhead in edge environments. Additionally, a dynamic adjustment mechanism evaluates augmentation success rates in real time, refining the selection processes to maintain model robustness. Experiments were conducted on two real-world datasets (TraceLog and FlowGraph) under simulated edge scenarios. Results show that HGAA consistently outperforms competitive baseline methods. Specifically, compared with the best non-adaptive augmentation strategies, HGAA achieves an average improvement of 4.5% in AUC and 4.6% in AP. Even larger gains are observed in challenging cases: for example, when using the HGT model on the TraceLog dataset, AUC improves by 14.6% and AP by 18.1%. Beyond accuracy, HGAA also significantly enhances efficiency: compared with filter-based methods, training time is reduced by up to 71% on TraceLog and 8.6% on FlowGraph, confirming its suitability for resource-constrained edge environments. These results highlight the potential of adaptive, edge-aware augmentation techniques in improving microservice anomaly detection within heterogeneous, resource-limited environments. Full article

With the rapid development of social media, short-text data have become increasingly important in fields such as public opinion monitoring, user feedback analysis, and intelligent recommendation systems. However, existing short-text sentiment analysis models often suffer from limited cross-domain adaptability and poor generalization performance. To address these challenges, this study proposes a novel short-text sentiment classification model based on the Bidirectional Encoder Representations from Transformers (BERTs) and a dual-stream Transformer gated attention mechanism. This model first employs Bidirectional Encoder Representations from Transformers (BERTs) and the Chinese Robustly Optimized BERT Pretraining Approach (Chinese-RoBERTa) to achieve data augmentation and multilevel semantic mining, thereby expanding the training corpus and enhancing minority class coverage. Second, a dual-stream Transformer gated attention mechanism was developed to dynamically adjust feature fusion weights, enhancing adaptability to heterogeneous texts. Finally, the model integrates a Bidirectional Gated Recurrent Unit (BiGRU) with Multi-Head Self-Attention (MHSA) to strengthen sequence information modeling and global context capture, enabling the precise identification of key sentiment dependencies. The model’s superior performance in handling data imbalance and complex textual sentiment logic scenarios is demonstrated by the experimental results, achieving significant improvements in accuracy and F1 score. The F1 score reached 92.4%, representing an average increase of 8.7% over the baseline models. This provides an effective solution for enhancing the performance and expanding the application scenarios of short-text sentiment analysis models. Full article

Assessing teaching behavior is essential for improving instructional quality, particularly in Physical Education, where classroom interactions are fast-paced and complex. Traditional evaluation methods such as questionnaires, expert observations, and manual discourse analysis are often limited by subjectivity, high labor costs, and poor scalability. These challenges underscore the need for automated, objective tools to support pedagogical assessment. This study explores and compares the use of Transformer-based language models for the automatic classification of teaching behaviors from real classroom transcriptions. A dataset of over 1300 utterances was compiled and annotated according to the teaching styles proposed in the circumplex approach (Autonomy Support, Structure, Control, and Chaos), along with an additional category for messages in which no style could be identified (Unidentified Style). To address class imbalance and enhance linguistic variability, data augmentation techniques were applied. Eight pretrained BERT-based Transformer architectures were evaluated, including several pretraining strategies and architectural structures. BETO achieved the highest performance, with an accuracy of 0.78, a macro-averaged F1-score of 0.72, and a weighted F1-score of 0.77. It showed strength in identifying challenging utterances labeled as Chaos and Autonomy Support. Furthermore, other BERT-based models purely trained with a Spanish text corpus like DistilBERT also present competitive performance, achieving accuracy metrics over 0.73 and and F1-score of 0.68. These results demonstrate the potential of leveraging Transformer-based models for objective and scalable teacher behavior classification. The findings support the feasibility of leveraging pretrained language models to develop scalable, AI-driven systems for classroom behavior classification and pedagogical feedback. Full article

This study presents the development and evaluation of an Augmented Reality (AR)-based training system aimed at improving patient setup accuracy in radiation therapy. Leveraging Microsoft HoloLens 2, the system provides an immersive environment for medical staff to enhance their understanding of patient setup procedures. High-resolution 3D anatomical models were reconstructed from CT scans using 3D Slicer, while Luma AI was employed to rapidly capture complete body surface models. Due to limitations in each method—such as missing extremities or back surfaces—Blender was used to merge the models, improving completeness and anatomical fidelity. The AR application was developed in Unity, employing spatial anchors and 125 × 125 mm² QR code markers to stabilize and align virtual models in real space. System accuracy testing demonstrated that QR code tracking achieved millimeter-level variation, with an expanded uncertainty of ±2.74 mm. Training trials for setup showed larger deviations in the X (left–right), Y (up-down), and Z (front-back) axes at the centimeter scale. This meant that we were able to quantify the user’s patient setup skills. While QR code positioning was relatively stable, manual placement of markers and the absence of real-time verification contributed to these errors. The system offers a radiation-free and interactive platform for training, enhancing spatial awareness and procedural skills. Future work will focus on improving tracking stability, optimizing the workflow, and integrating real-time feedback to move toward clinical applicability. Full article