3031 papers
This paper proposes an Auxiliary Classifier Generative Adversarial Network (AC-GAN) framework that synthesizes high-fidelity thermal images of rare photovoltaic defects to address class imbalance, thereby significantly improving the probabilistic calibration and reliability of AI-driven fault diagnosis in solar operations.
This paper introduces QNeuroGen, a hybrid quantum-classical neural framework that achieves real-time cognitive state decoding from noisy EEG signals with 89.2% accuracy and 48 ms latency by integrating structured preprocessing with variational quantum feature encoding.
This study proposes a hybrid ResNet101-LSTM architecture optimized with Adam that achieves high accuracy and specificity in classifying breast cancer from BreaKHis histopathological images at 40x magnification, outperforming other deep learning models in both binary and multiclass scenarios.
This study demonstrates the viability of Quantum Neural Networks for wind energy forecasting by systematically evaluating 12 configurations against classical models, revealing that specific QNN designs can achieve competitive performance (up to R² = 0.94) despite the computational trade-offs inherent in NISQ-era simulations.
This paper introduces QGH-256, a novel quantum hashing algorithm that generates 256-bit fingerprints by mapping messages to weighted graphs on a toroidal grid and utilizing Quantum Phase Estimation to extract spectral features that capture both structural and dynamical properties, demonstrating strong cryptographic sensitivity and feasibility through Qiskit simulations.
This paper introduces DDANet, a novel Dense Dual-Attention Network that combines dense skip connections and dual-attention mechanisms to achieve superior boundary delineation and generalization in medical image segmentation, outperforming baseline models with high accuracy, Dice, and IoU scores.
This paper presents TinyOL-ARLDA, a lightweight, on-device incremental learning framework based on regularized Linear Discriminant Analysis that enables real-time, cloud-independent sound-based predictive maintenance on microcontrollers with over 94% accuracy and a memory footprint under 20 KB.
This paper proposes CGS-RAG, a novel framework that leverages the HDRF algorithm for customizable community partitioning and semantic reporting to overcome the limitations of traditional and graph-enhanced RAG systems by balancing global semantic consistency with efficient resource usage in dynamic knowledge scenarios.
This paper presents a deterministic, concept-grounded framework for flight simulator training that maps subsymbolic evidence to admissible scenario adaptations via a constrained machine learning policy, ensuring trustworthiness through formal safety guarantees and demonstrating empirical effectiveness in approximating a verified baseline policy.
This paper proposes a multi-modal computational framework that fuses gait video, inertial signals, plantar pressure, and knee flexion data to achieve high-accuracy intelligent assessment of knee function and exercise risk prediction for elderly patients, demonstrating superior stability and robustness compared to existing baseline models.