Search Results (286)

Autonomous racing serves as a challenging testbed that exposes the limitations of perception-decision-control algorithms in extreme high-speed environments, revealing safety gaps not addressed in existing autonomous driving research. However, traditional control techniques (e.g., FGM and MPC) and reinforcement learning-based approaches (including model-free and Dreamer variants) struggle to simultaneously satisfy sample efficiency, prediction reliability, and real-time control performance, making them difficult to apply in actual high-speed racing environments. To address these challenges, we propose LiDAR Dreamer, a novel world model specialized for LiDAR sensor data. LiDAR Dreamer introduces three core techniques: (1) efficient point cloud preprocessing and encoding via Cartesian Polar Bar Charts, (2) Light Structured State-Space Cells (LS3C) that reduce RSSM parameters by 14.2% while preserving key dynamic information, and (3) a Displacement Covariance Distance divergence function, which enhances both learning stability and expressiveness. Experiments in PyBullet F1TENTH simulation environments demonstrate that LiDAR Dreamer achieves competitive performance across different track complexities. On the Austria track with complex corners, it reaches 90% of DreamerV3’s performance (1.14 vs. 1.27 progress) while using 81.7% fewer parameters. On the simpler Columbia track, while model-free methods achieve higher absolute performance, LiDAR Dreamer shows improved sample efficiency compared to baseline Dreamer models, converging faster to stable performance. The Treitlstrasse environment results demonstrate comparable performance to baseline methods. Furthermore, beyond the 14.2% RSSM parameter reduction, reward loss converged more stably without spikes, improving overall training efficiency and stability. Full article

To address the issue of reduced image contrast and visibility caused by turbid media, such as dense fog, this paper proposes a novel polarization-based single-image dehazing model. The model introduces a first-of-its-kind nonlinear joint polarization model for airlight and target light. This model is established within a Cartesian coordinate system, abstracted as an analytical geometric model. Leveraging the structural similarity principle in images, boundary constraints are applied to enhance the accuracy of target light estimation. Finally, image dehazing and enhancement are achieved using the atmospheric scattering model. Experimental results demonstrate that the proposed algorithm does not rely on dataset training, maintains the highest structural consistency, and achieves superior image restoration across various scenarios, producing results that most closely resemble natural observation. Full article

Rice is one of the most important staple crops in China, and the rapid and accurate extraction of rice planting areas plays a crucial role in the agricultural management and food security assessment. However, the existing rice field identification methods faced the significant challenges in mountainous regions due to the severe cloud contamination, insufficient utilization of multi-dimensional features, and limited classification accuracy. This study presented a novel rice field identification method based on the Graph Convolutional Networks (GCN) that effectively integrated multi-source remote sensing data tailored for the complex mountainous terrain. A coarse-to-fine cloud removal strategy was developed by fusing the synthetic aperture radar (SAR) imagery with temporally adjacent optical remote sensing imagery, achieving high cloud removal accuracy, thereby providing reliable and clear optical data for the subsequent rice mapping. A comprehensive multi-feature library comprising spectral, texture, polarization, and terrain attributes was constructed and optimized via a stepwise selection process. Furthermore, the 19 key features were established to enhance the classification performance. The proposed method achieved an overall accuracy of 98.3% for the rice field identification in Huoshan County of the Dabie Mountains, and a 96.8% consistency compared to statistical yearbook data. The ablation experiments demonstrated that incorporating terrain features substantially improved the rice field identification accuracy under the complex topographic conditions. The comparative evaluations against support vector machine (SVM), random forest (RF), and U-Net models confirmed the superiority of the proposed method in terms of accuracy, local performance, terrain adaptability, training sample requirement, and computational cost, and demonstrated its effectiveness and applicability for the high-precision rice field distribution mapping in mountainous environments. Full article

Spiking neural networks (SNNs) have emerged as a promising model for energy-efficient, event-driven processing of asynchronous event streams from Dynamic Vision Sensors (DVSs), a class of neuromorphic image sensors with microsecond-level latency and high dynamic range. Nevertheless, challenges persist in optimising training and effectively handling spatio-temporal complexity, which limits their potential for real-time applications on embedded sensing systems such as object tracking and recognition. Targeting this neuromorphic sensing pipeline, this paper proposes the Dynamic Tracking with Event Attention Spiking Network (DTEASN), a novel framework designed to address these challenges by employing a pure SNN architecture, bypassing conventional convolutional neural network (CNN) operations, and reducing GPU resource dependency, while tailoring the processing to DVS signal characteristics (asynchrony, sparsity, and polarity). The model incorporates two innovative, self-developed components: an event-driven multi-scale attention mechanism and a spatio-temporal event convolver, both of which significantly enhance spatio-temporal feature extraction from raw DVS events. An Event-Weighted Spiking Loss (EW-SLoss) is introduced to optimise the learning process by prioritising informative events and improving robustness to sensor noise. Additionally, a lightweight event tracking mechanism and a custom synaptic connection rule are proposed to further improve model efficiency for low-power, edge deployment. The efficacy of DTEASN is demonstrated through empirical results on event-based (DVS) object recognition and tracking benchmarks, where it outperforms conventional methods in accuracy, latency, event throughput (events/s) and spike rate (spikes/s), memory footprint, spike-efficiency (energy proxy), and overall computational efficiency under typical DVS settings. By virtue of its event-aligned, sparse computation, the framework is amenable to highly parallel neuromorphic hardware, supporting on- or near-sensor inference for embedded applications. Full article

The Arctic, characterized by extensive ice and snow cover with persistent low solar elevation angles and prolonged polar nights, poses significant challenges for conventional spectral threshold methods in cloud detection and cloud-top thermodynamic phase classification. The study addressed these limitations by combining active and passive remote sensing and developing a machine learning framework for cloud detection and cloud-top thermodynamic phase classification. Utilizing the CALIOP (Cloud-Aerosol Lidar with Orthogonal Polarization) cloud product from 2021 as the truth reference, the model was trained with spatiotemporally collocated datasets from FY3D/MERSI-II (Medium Resolution Spectral Imager-II) and CALIOP. The AdaBoost (Adaptive Boosting) machine learning algorithm was employed to construct the model, with considerations for six distinct Arctic surface types to enhance its performance. The accuracy test results showed that the cloud detection model achieved an accuracy of 0.92, and the cloud recognition model achieved an accuracy of 0.93. The inversion performance of the final model was then rigorously evaluated using a completely independent dataset collected in July 2022. Our findings demonstrated that our model results align well with results from CALIOP, and the detection and identification outcomes across various surface scenarios show high consistency with the actual situations displayed in false-color images. Full article

Maritime transportation, fishing, scientific research, and other activities rely on various types of ships and platforms, making precise monitoring of ships at sea essential. Synthetic Aperture Radar (SAR) is minimally affected by weather conditions and darkness and is used for ship detection in maritime environments. This study analyzes the differences in backscatter characteristics among various ship types in SAR images and proposes SISS-YOLO, an enhanced model based on YOLOv8. The proposed method addresses the challenge of ship instance segmentation in SAR images involving multiple polarizations, scenarios, and classes. First, the backbone structure was optimized by incorporating additional pooling layers and refining the activation functions. Second, the Coordinate Attention (CA) module was integrated into the C2F template, embedding spatial position information into the channel attention mechanism. Third, a slide loss function was adopted to address the class imbalance across ship categories. The experiments were conducted on the OpenSARShip2.0 dataset, which includes cargo, tanker, passenger and engineering ships. The results show that the SISS-YOLO achieves a mask precision of 88.3%, a mask recall of 86.4% and a mask mAP50 of 93.4% for engineering ships. Compared with YOLOv8m, SISS-YOLO achieved improvements of 15.7% in mask precision and 8.8% in mask recall. The model trained on the OpenSARShip2.0 dataset was directly applied to the FUSAR-Ship1.0 dataset, demonstrating a degree of robustness. When applied to SAR data, the SISS-YOLO model achieves high detection accuracy, demonstrating generalization. Full article

Accurate and non-invasive assessment of fat and muscle composition is crucial for biomedical monitoring to track health conditions in humans and pets, as well as for classifying meats in the meat industry. This study introduces a cost-effective, multifunctional ultra-wideband microwave system operating from 2.4 to 4.4 GHz, designed for rapid and non-destructive quantification of fat thickness, muscle thickness, and fat-to-muscle ratio in diverse ex vivo samples, including pork, beef, and oil–water mixtures. The compact handheld device integrates essential RF components such as a frequency synthesizer, directional coupler, logarithmic power detector, and a dual-polarized Vivaldi antenna. Bluetooth telemetry enables seamless real-time data transmission to mobile- or PC-based platforms, with each measurement completed in a few seconds. To enhance signal quality, a two-stage denoising pipeline combining low-pass filtering and Savitzky–Golay smoothing was applied, effectively suppressing noise while preserving key spectral features. Using a random forest regression model trained on resonance frequency and signal-loss features, the system demonstrates high predictive performance even under limited sample conditions. Correlation coefficients for fat thickness, muscle thickness, and fat-to-muscle ratio consistently exceeded 0.90 across all sample types, while mean absolute errors remained below 3.5 mm. The highest prediction accuracy was achieved in homogeneous oil–water samples, whereas biologically complex tissues like pork and beef introduced greater variability, particularly in muscle-related measurements. The proposed microwave system is highlighted as a highly portable and time-efficient solution, with measurements completed within seconds. Its low cost, ability to analyze multiple tissue types using a single device, and non-invasive nature without the need for sample pre-treatment or anesthesia make it well suited for applications in agri-food quality control, point-of-care diagnostics, and broader biomedical fields. Full article

Ionospheric scintillation represents a disturbance phenomenon induced by irregular electron density variations, predominantly occurring in equatorial, auroral, and polar regions, thereby posing significant threats to Global Navigation Satellite Systems (GNSS) performance. Polar regions in particular confront distinctive challenges, including the sparse deployment of dedicated ionospheric scintillation monitoring receiver (ISMR) equipment, the limited availability of strong scintillation samples, severely imbalanced training datasets, and the insufficient sensitivity of conventional Deep Neural Networks (DNNs) to intense scintillation events. To address these challenges, this study proposes a modeling framework that integrates residual neural networks (ResNet) with the Synthetic Minority Over-sampling Technique for Regression with Gaussian Noise (SMOGN). The proposed model incorporates multi-source disturbance features to accurately estimate phase scintillation indices (${\sigma }_{\phi }$) in polar regions. The methodology was implemented and validated across multiple polar observation stations in Canada. Shapley Additive Explanations (SHAP) interpretability analysis reveals that the rate of total electron content index (ROTI) features contribute up to 64.09% of the predictive weight. The experimental results demonstrate a substantial performance enhancement compared with conventional DNN models, with root mean square error (RMSE) values ranging from 0.0078 to 0.038 for daytime samples in 2024, and an average coefficient of determination (R²) consistently exceeding 0.89. The coefficient of determination for the Pseudo-Random Noise (PRN) path estimation results can reach 0.91. The model has good estimation results at different latitudes and is able to accurately capture the distribution characteristics of the local strong scintillation structures and their evolution patterns. Full article

Understanding and modeling the flow of information in human societies is essential for capturing phenomena such as polarization, opinion formation, and misinformation diffusion. Traditional agent-based models often rely on simplified behavioral rules that fail to capture the nuanced and context-sensitive nature of human decision-making. In this study, we explore the potential of Large Language Models (LLMs) as data-driven, high-fidelity agents capable of simulating individual opinions under varying informational conditions. Conditioning LLMs on real survey data from the 2020 American National Election Studies (ANES), we investigate their ability to predict individual-level responses across a spectrum of political and social issues in a zero-shot setting, without any training on the survey outcomes. Using Jensen–Shannon distance to quantify divergence in opinion distributions and F1-score to measure predictive accuracy, we compare LLM-generated simulations to those produced by a supervised Random Forest model. While performance at the individual level is comparable, LLMs consistently produce aggregate opinion distributions closer to the empirical ground truth. These findings suggest that LLMs offer a promising new method for simulating complex opinion dynamics and modeling the probabilistic structure of belief systems in computational social science. Full article

User-generated content, such as product and app reviews, offers more than just sentiment. It provides a rich spectrum of emotional expression that reveals users’ experiences, frustrations and expectations. Traditional sentiment analysis, which typically classifies text as positive or negative, lacks the nuance needed to fully understand the emotional drivers behind customer feedback. In this research, we focus on fine-grained emotion classification using core emotions. By identifying specific emotions rather than sentiment polarity, we enable more actionable insights for e-commerce and app development, supporting strategies such as feature refinement, marketing personalization and proactive customer engagement. We leverage the Hugging Face Emotions dataset and adopt a two-phase modeling approach. In the first phase, we use a pre-trained DistilBERT model as a feature extractor and evaluate multiple classical classifiers (Logistic Regression, Support Vector Classifier, Random Forest) to establish performance baselines. In the second phase, we fine-tune the DistilBERT model end-to-end using the Hugging Face Trainer API, optimizing classification performance through task-specific adaptation. Training is tracked using the Weights & Biases (wandb) API. Comparative analysis highlights the substantial performance gains from fine-tuning, particularly in capturing informal or noisy language typical in user reviews. The final fine-tuned model is applied to a dataset of customers’ reviews, identifying the dominant emotions expressed. Our results demonstrate the practical value of emotion-aware analytics in uncovering the underlying “why” behind user sentiment, enabling more empathetic decision-making across product design, customer support and user experience (UX) strategy. Full article

Mueller matrix polarimetry is recently attracting more and more attention for its diagnostic potentials. However, for prevalently used division of time Mueller matrix polarimeter based on dual-rotating retarder scheme, beam drift induced by rotating polarizers and waveplates introduces spatial misalignment and pseudo-edge artifacts in imaging results, hindering following accurate microstructural features characterization. In this paper, we quantitatively analyze the beam drift phenomenon in dual-rotating retarder Mueller matrix microscopy and its impact on linear retardance measurement, which is frequently used to reflect tissue fiber arrangement. It is demonstrated that polarizer rotation induces larger beam drift than waveplate rotation due to surface non-uniformity and stress deformation. Furthermore, for waveplates rotated constantly in dual-rotating retarder scheme, their tilt within polarization state analyzer can result in more drift and throughput loss than those within polarization state generator. Finally, phantom and tissue experiments confirm that beam drift, rather than inherent optical path changes, dominates the systematic overestimation of linear retardance in boundary image regions. The findings highlight beam drift as a dominant error source for quantifying linear retardance, necessitating careful optical design alignment and a reliable registration algorithm to obtain highly accurate polarization data for training machine learning models of pathological diagnosis using Mueller matrix microscopy. Full article

Photocatalysis is an emerging technology that harnesses light energy to facilitate chemical reactions. It has garnered considerable attention in the field of catalysis due to its promising applications in environmental remediation and sustainable energy generation. Recently, researchers have been exploring innovative techniques to improve the surface reactivity of ferroelectric materials for catalytic purposes, leveraging their distinct properties to enhance photocatalytic efficiency. With their switchable polarization and improved charge transport capabilities, ferroelectric materials show promise as effective photocatalysts for various reactions, including carbon dioxide (CO₂) reduction. Through a blend of experimental studies and theoretical modeling, researchers have shown that these materials can effectively convert CO₂ into valuable products, contributing to efforts to reduce greenhouse gas emissions and promote a cleaner environment. An artificial neural network (ANN) was employed to analyze parameter relationships and their impacts in this study, demonstrating its ability to manage training data errors and its applications in fields like speech and image recognition. This research also examined changes in charge separation, light absorption, and surface area related to variations in band gap and polarization, confirming prediction accuracy through linear regression analysis. Full article

This study addresses the shortcomings of traditional exercise volume assessment methods in dynamic modeling and individual adaptation by proposing a multi-modal data fusion framework based on a spatio-temporal attention-enhanced LSTM neural network for personalized exercise volume optimization in college basketball courses. By integrating physiological signals (heart rate), kinematic parameters (triaxial acceleration, step count), and environmental data collected from smart wearable devices, we constructed a dynamic weighted fusion mechanism and a personalized correction engine, establishing an evaluation model incorporating BMI correction factors and fitness-level compensation. Experimental data from 100 collegiate basketball trainees (60 males, 40 females; BMI 17.5–28.7) wearing Polar H10 and Xsens MVN devices were analyzed through an 8-week longitudinal study design. The framework integrates physiological monitoring (HR, HRV), kinematic analysis (3D acceleration at 100 Hz), and environmental sensing (SHT35 sensor). Experimental results demonstrate the following: (1) the LSTM-attention model achieves 85.3% accuracy in exercise intensity classification, outperforming traditional methods by 13.2%, with its spatio-temporal attention mechanism effectively capturing high-dynamic movement features such as basketball sudden stops and directional changes; (2) multi-modal data fusion reduces assessment errors by 15.2%, confirming the complementary value of heart rate and acceleration data; (3) the personalized correction mechanism significantly improves evaluation precision for overweight students (error reduction of 13.6%) and beginners (recognition rate increase of 18.5%). System implementation enhances exercise goal completion rates by 10.3% and increases moderate-to-vigorous training duration by 14.7%, providing a closed-loop “assessment-correction-feedback” solution for intelligent sports education. The research contributes methodological innovations in personalized modeling for exercise science and multi-modal time-series data processing. Full article

Atomic magnetometers (AMs), recognized for their ultra-high magnetic sensitivity, demand highly uniform pump light fields to maximize measurement accuracy. In this paper, a phase modulation-based method using convolutional neural networks (CNN) and the Gerchberg–Saxton (GS) algorithm is proposed to generate the pumping light field, and the model was trained using a supervised learning approach with a custom dataset. The specific training settings are as follows: the backpropagation algorithm was adopted as the training algorithm, and the Adam optimization method was used for network training, with a learning rate of 0.001 and a total of 100 training epochs, utilizing a liquid crystal spatial light modulator (LCSLM) to regulate the light field phase distribution dynamically. By transforming Gaussian beams into flat-top beams, the method significantly enhances polarization uniformity within vapor cells, leading to improved magnetometric sensitivity. The proposed hybrid algorithm reduces the mean square error from 35% to 19% and peak non-uniformity from 21% to 7.6%. A reflective LCSLM-based optical setup is implemented to produce circular and square flat-top beams with a measured non-uniformity of 5.1%, resulting in an enhancement of magnetic sensitivity from 14.54 fT/Hz^1/2 to 7.80 fT/Hz^1/2. Full article

In clinical decision support systems (CDSSs), where accurate classification of drug–drug interactions (DDIs) can directly affect treatment safety and outcomes, identifying drug interactions is a major challenge, introducing a scalable approach for classifying DDIs utilizing a finely-tuned biomedical language model. The method shown here uses BiomedBERT, a domain-specific version of bidirectional encoder representations from transformers (BERT) that was pre-trained on biomedical literature, to reduce the number of resources needed during fine-tuning. Low-rank adaptation (LoRA) was used to fine-tune the model on the DrugBank dataset. The objective was to classify DDIs into two clinically distinct categories, that is, synergistic and antagonistic interactions. A pseudo-labeling strategy was created to deal with the problem of not having enough labeled data. A curated ground-truth dataset was constructed using polarity-labeled interaction entries from DrugComb and verified DrugBank antagonism pairs. The fine-tuned model is used to figure out what kinds of interactions there are in the rest of the unlabeled data. A checkpointing system saves predictions and confidence scores in small pieces, which means that the process can be continued and is not affected by system crashes. The framework is designed to log every prediction it makes, allowing results to be refined later, either manually or through automated updates, without discarding low-confidence cases, as traditional threshold-based methods often do. The method keeps a record of every output it generates, making it easier to revisit earlier predictions, either by experts or with improved tools, without depending on preset confidence cutoffs. It was built with efficiency in mind, so it can handle large amounts of biomedical text without heavy computational demands. Rather than focusing on model novelty, this research demonstrates how existing biomedical transformers can be adapted to polarity-aware DDI classification with minimal computational overhead, emphasizing deployment feasibility and clinical relevance. Full article