Search Results (794)

3D point cloud object tracking plays a critical role in fields such as autonomous driving and robotics, making the security of these models essential. Adversarial attacks are a key approach for studying the robustness and security of tracking models. However, research on the generalization of adversarial attacks for 3D point-cloud-tracking models is limited, and the frequency-domain information of the point cloud’s geometric structure is often overlooked. This frequency information is closely related to the generalization of 3D point-cloud-tracking models. To address these limitations, this paper proposes a novel adversarial method for 3D point cloud object tracking, utilizing frequency-domain attacks based on the importance of frequency bands. The attack operates in the frequency domain, targeting the low-frequency components of the point cloud within the search area. To make the attack more targeted, the paper introduces a frequency band importance saliency map, which reflects the significance of sub-frequency bands for tracking and uses this importance as attack weights to enhance the attack’s effectiveness. The proposed attack method was evaluated on mainstream 3D point-cloud-tracking models, and the adversarial examples generated from white-box attacks were transferred to other black-box tracking models. Experiments show that the proposed attack method reduces both the average success rate and precision of tracking, proving the effectiveness of the proposed adversarial attack. Furthermore, when the white-box adversarial samples were transferred to the black-box model, the tracking metrics also decreased, verifying the transferability of the attack method. Full article

Cloud cover can delay landslide detection in optical satellite imagery for weeks, complicating disaster response. Synthetic Aperture Radar (SAR) backscatter imagery, which is widely used for monitoring floods and avalanches, remains underutilised for landslide detection due to a limited understanding of landslide signatures in SAR data. We developed a conceptual model of landslide expression in SAR backscatter (σ°) change images through iterative investigation of over 1000 landslides across 30 diverse study areas. Using multi-temporal composites and dense time series Sentinel-1 C-band SAR data, we identified characteristic patterns linked to land cover, terrain, and landslide material. The results showed either increased or decreased backscatter depending on environmental conditions, with reduced visibility in urban or mixed vegetation areas. Detection was also hindered by geometric distortions and snow cover. The diversity of landslide expression illustrates the need to consider local variability and multi-track (ascending and descending) satellite data in designing representative training datasets for automated detection models. The conceptual model was applied to three recent disaster events using the first post-event Sentinel-1 image, successfully identifying previously unknown landslides before optical imagery became available in two cases. This study provides a theoretical foundation for interpreting landslides in SAR imagery and demonstrates its utility for rapid landslide detection. The findings support further exploration of rapid landslides in SAR backscatter data and future development of automated detection models, offering a valuable tool for disaster response. Full article

This paper introduces CodeDive, a web-based programming environment with real-time behavioral tracking designed to enhance student progress assessment and provide timely support for learners, while also addressing the academic integrity challenges posed by Large Language Models (LLMs). Visibility into the student’s learning process has become essential for effective pedagogical analysis and personalized feedback, especially in the era where LLMs can generate complete solutions, making it difficult to truly assess student learning and ensure academic integrity based solely on the final outcome. CodeDive provides this process-level transparency by capturing fine-grained events, such as code edits, executions, and pauses, enabling instructors to gain actionable insights for timely student support, analyze learning trajectories, and effectively uphold academic integrity. It operates on a scalable Kubernetes-based cloud architecture, ensuring security and user isolation via containerization and SSO authentication. As a browser-accessible platform, it requires no local installation, simplifying deployment. The system produces a rich data stream of all interaction events for pedagogical analysis. In a Spring 2025 deployment in an Operating Systems course with approximately 100 students, CodeDive captured nearly 25,000 code snapshots and over 4000 execution events with a low overhead. The collected data powered an interactive dashboard visualizing each learner’s coding timeline, offering actionable insights for timely student support and a deeper understanding of their problem-solving strategies. By shifting evaluation from the final artifact to the developmental process, CodeDive offers a practical solution for comprehensively assessing student progress and verifying authentic learning in the LLM era. The successful deployment confirms that CodeDive is a stable and valuable tool for maintaining pedagogical transparency and integrity in modern classrooms. Full article

The evolution of the winemaking industry towards intelligent and digitalized systems is crucial for precision winemaking and ensuring product safety. In this context, the Internet of Things (IoT) provides a key strategy for real-time monitoring and data management throughout the winemaking process. However, comprehensive multi-parameter IoT-based monitoring and time-series prediction of physicochemical parameters during storage are currently lacking, limiting the ability to assess storage conditions and provide early warning of quality deterioration. To address these gaps, a multi-parameter IoT monitoring system was designed and developed to track conductivity, dissolved oxygen, and temperature in real time. Data were transmitted via a 4th-generation (4G) mobile communication module to the TLINK cloud platform for storage and visualization. An 80-day storage experiment confirmed the system’s reliability for long-term monitoring, and analysis of parameter trends demonstrated its effectiveness in assessing storage conditions and wine quality evolution. Furthermore, Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Temporal Convolutional Network (TCN) models, and Autoregressive Integrated Moving Average (ARIMA) were implemented to predict physicochemical parameter trends. The TCN model achieved the highest predictive performance, with coefficients of determination (R²) of 0.955, 0.968, and 0.971 for conductivity, dissolved oxygen, and temperature, respectively, while LSTM and GRU showed comparable results. These results demonstrate that integrating IoT-based multi-parameter monitoring with deep learning time-series prediction enables real-time detection of abnormal storage and quality deterioration, providing a novel and practical framework for early warning throughout the wine storage process. Full article

The NASA CERES project conducts an independent assessment of the calibration stability of MODIS and VIIRS reflective solar bands to ensure consistency in CERES-derived clouds and radiative flux products. The assessment includes the use of tropical deep convective cloud invariant targets (DCC-IT), identified using a simple brightness temperature threshold. For visible bands, the collective DCC pixel radiance probability density function (PDF) was negatively skewed. By tracking the bright inflection point, rather than the PDF mode, and applying an anisotropic adjustment suited for the brightest DCC radiances, the lowest trend standard errors were obtained within 0.26% for NPP-VIIRS and within 0.36% for NOAA20-VIIRS and Aqua-MODIS. A kernel density estimation function was used to infer the PDF, which avoided discretization noise caused by sparse sampling. The near 10° regional consistency of the anisotropic corrected PDF inflection point radiances validated the DCC-IT approach. For the shortwave infrared (SWIR) bands, the DCC radiance variability is dependent on the ice particle scattering and absorption and is band-specific. The DCC radiance varies regionally, diurnally, and seasonally; however, the inter-annual variability is much smaller. Empirical bidirectional reflectance distribution functions (BRDFs), constructed from multi-year records, were most effective in characterizing the anisotropic behavior. Due to the distinct land and ocean as well as regional radiance differences, land, ocean, and regional BRDFs were evaluated. The regional radiance variability was mitigated by normalizing the individual regional radiances to the tropical mean radiance. Because the DCC pixel radiances have a Gaussian distribution, the mean radiance was used to track the DCC response. The regional BRDF-adjusted DCC-IT mean radiance trend standard errors were within 0.38%, 0.46%, and 1% for NOAA20-VIIRS, NPP-VIIRS, and Aqua-MODIS, respectively. Full article

Forest fire risk is rising with climate warming, highlighting the need for timely monitoring and early warning. Satellite-based monitoring, currently a primary tool in remote sensing for fire detection, suffers from spatiotemporal gaps due to limited resolution and cloud cover. This study developed a novel smoke detection technology using operational S-band dual-polarization weather radar. By analyzing six forest fire cases in Zhejiang Province, China (2023), we established a filtering method using dual-polarization parameters, with thresholds set to a differential reflectivity (Z_DR) ≥ 3 dB and a cross-correlation coefficient (${\rho }_{HV}$) ≤ 0.7. This method effectively isolates fire-related echoes and, compared with geostationary satellites, enables more continuous monitoring; it also detects small and early-stage fires. Furthermore, radar-derived fire perimeters closely match satellite imagery, demonstrating its potential for real-time fire-spread tracking. The high spatiotemporal resolution and multi-parameter advantages of dual-polarization radar can complement satellite observations, offering vital support for early warning and real-time decision-making in fire management. Full article

In aluminum electrolysis workshops, real-time pose perception of shelling heads is crucial to process accuracy and equipment safety. However, due to high temperatures, smoke, dust, and metal obstructions, traditional pose estimation methods struggle to achieve high accuracy and robustness. At the same time, the continuous movement of the shelling head and the similar geometric structures around it make it hard to match point-clouds, which makes it even harder to track the position and orientation. In response to the above challenges, we propose a multi-stage optimization pose estimation algorithm based on point-cloud processing. This method is designed for dynamic perception tasks of three-dimensional components in complex industrial scenarios. First stage improves the accuracy of initial matching by combining a weighted 3D Hough voting and adaptive threshold mechanism with an improved FPFH feature matching strategy. In the second stage, by integrating FPFH and PCA feature information, a stable initial registration is achieved using the RANSAC-IA coarse registration framework. In the third stage, we designed an improved ICP algorithm that effectively improved the convergence of the registration process and the accuracy of the final pose estimation. The experimental results show that the proposed method has good robustness and adaptability in a real electrolysis workshop environment, and can achieve pose estimation of the shelling head in the presence of noise, occlusion, and complex background interference. Full article

The growing demand for electricity and the urgent need to reduce environmental impact have made sustainable energy utilization a global priority. Street lighting, as a significant consumer of urban electricity, requires innovative solutions to enhance efficiency and reliability. This study presents an off-grid smart street lighting system that combines solar photovoltaic generation with battery storage and Internet of Things (IoT)-based control to ensure continuous and efficient operation. The system integrates Long Range Wide Area Network (LoRaWAN) communication technology for remote monitoring and control without internet connectivity and employs the Perturb and Observe (P&O) maximum power point tracking (MPPT) algorithm to maximize energy extraction from solar sources. Data transmission from the LoRaWAN gateway to the cloud is facilitated through the Message Queuing Telemetry Transport (MQTT) protocol, enabling real-time access and management via a graphical user interface. Experimental results demonstrate that the proposed system achieves a maximum MPPT efficiency of 97.96%, supports reliable communication over distances of up to 10 km, and successfully operates four LED streetlights, each spaced 400 m apart, across an open area of approximately 1.2 km—delivering a practical, energy-efficient, and internet-independent solution for smart urban infrastructure. Full article

Background/Objectives: Chronic kidney disease (CKD) requires intensive dietary and lifestyle management, yet patient engagement and access to tailored education remain limited, particularly outside clinical settings. This study describes the development and implementation of NephroGo, and evaluates its usability, user engagement, and perceived acceptability among patients with CKD. Methods: The app was developed based on clinical and dietary guidelines, incorporating personalized nutrient recommendations, dialysis tracking, and educational content. Technically, it features a Django backend, Flutter mobile frontend, and secure cloud-based hosting. User feedback was collected through one-time interviews (n = 10) and a standardized Mobile App Rating Scale (MARS) survey (n = 32). Longitudinal usage data over four years were also analyzed. Results: Initially, NephroGo was downloaded by 204 users, of whom 93.6% were considered active users based on defined behavioral engagement thresholds. Over a four-year period, the app accumulated a total of 1670 downloads. This study focuses on evaluating user engagement, usability, and perceived acceptability of the NephroGo app over a four-year period. Most users were female (52.3%) and aged 30–65. Stage 5 CKD patients and those undergoing peritoneal dialysis (PD) had the highest engagement. The most-used feature was the personalized nutrition calculator, with sodium being the most frequently exceeded nutrient. The average MARS score was 4.09 ± 0.66, with functionality rated highest (4.27 ± 0.74). App ratings were significantly higher among users referred by physicians (p = 0.039). Conclusions: NephroGo offers a scalable digital tool to support dietary management and health monitoring, with potential to complement standard nephrology care in a resource-conscious manner. Full article

Calibration of power equipment has become an essential task in modern power systems. This paper proposes a distributed remote calibration prototype based on a cloud–edge–end architecture by integrating intelligent sensing, Internet of Things (IoT) communication, and edge computing technologies. The prototype employs a high-precision frequency-to-voltage conversion module leveraging satellite signals to address traceability and value transmission challenges in remote calibration, thereby ensuring reliability and stability throughout the process. Additionally, an environmental monitoring module tracks parameters such as temperature, humidity, and electromagnetic interference. Combined with video surveillance and optical character recognition (OCR), this enables intelligent, end-to-end recording and automated data extraction during calibration. Furthermore, a cloud-edge task scheduling algorithm is implemented to offload computational tasks to edge nodes, maximizing resource utilization within the cloud–edge collaborative system and enhancing service quality. The proposed prototype extends existing cloud–edge collaboration frameworks by incorporating calibration instruments and sensing devices into the network, thereby improving the intelligence and accuracy of remote calibration across multiple layers. Furthermore, this approach facilitates synchronized communication and calibration operations across symmetrically deployed remote facilities and reference devices, providing solid technical support to ensure that measurement equipment meets the required precision and performance criteria. Full article

Accurate boundary detection is critical for autonomous trackless rubber-wheeled vehicles in underground coal mines, as it prevents lateral collisions with tunnel walls. Unlike open-road environments, underground tunnels suffer from poor illumination, water mist, and dust, which degrade visual imaging. To address these challenges, this paper proposes a navigable area computation for underground autonomous vehicles via ground feature and boundary processing, consisting of three core steps. First, a real-time point cloud correction process via pre-correction and dynamic update aligns ground point clouds with the LiDAR coordinate system to ensure parallelism. Second, corrected point clouds are projected onto a 2D grid map using a grid-based method, effectively mitigating the impact of ground unevenness on boundary extraction; third, an adaptive boundary completion method is designed to resolve boundary discontinuities in junctions and shunting chambers. Additionally, the method emphasizes continuous extraction of boundaries over extended periods by integrating temporal context, ensuring the continuity of boundary detection during vehicle operation. Experiments on real underground vehicle data validate that the method achieves accurate detection and consistent tracking of dual-sided boundaries across straight tunnels, curves, intersections, and shunting chambers, meeting the requirements of underground autonomous driving. This work provides a rule-based, real-time solution feasible under limited computing power, offering critical safety redundancy when deep learning methods fail in harsh underground environments. Full article

Complex event detection (CED) adds value to camera stream data in various applications such as workplace safety, task monitoring, security, and health. Recent CED frameworks have addressed the issues of limited spatiotemporal labels and costly training by decomposing the CED into low-level features, as well as spatial and temporal relationship extraction. However, these frameworks suffer from high resource costs, low scalability, and an increased number of false positives and false negatives. This paper proposes GICEDCAM, which distributes CED across edge, stateless, and stateful layers to improve scalability and reduce computation cost. Additionally, we introduce a Spatial Event Corrector component that leverages geospatial data analysis to minimize false negatives and false positives in spatial event detection. We evaluate GICEDCAM on 16 camera streams covering four complex events. Relative to a strong open-source baseline configured for our setting, GICEDCAM reduces end-to-end latency by 36% and total computational cost by 45%, with the advantage widening as objects per frame increase. Among corrector variants, Bayesian Network (BN) yields the lowest latency, Long Short-Term Memory (LSTM) achieves the highest accuracy, and trajectory analysis offers the best accuracy–latency trade-off for this architecture. Full article

Most existing Simultaneous Localization and Mapping (SLAM) systems rely on the assumption of static environments to achieve reliable and efficient mapping. However, such methods often suffer from degraded localization accuracy and mapping consistency in dynamic settings, as they lack explicit mechanisms to distinguish between static and dynamic elements. To overcome this limitation, we present BMP-SLAM, a vision-based SLAM approach that integrates semantic segmentation and Bayesian motion estimation to robustly handle dynamic indoor scenes. To enable real-time dynamic object detection, we integrate YOLOv5, a semantic segmentation network that identifies and localizes dynamic regions within the environment, into a dedicated dynamic target detection thread. Simultaneously, the data association Bayesian mobile probability proposed in this paper effectively eliminates dynamic feature points and successfully reduces the impact of dynamic targets in the environment on the SLAM system. To enhance complex indoor robotic navigation, the proposed system integrates semantic keyframe information with dynamic object detection outputs to reconstruct high-fidelity 3D point cloud maps of indoor environments. The evaluation conducted on the TUM RGB-D dataset indicates that the performance of BMP-SLAM is superior to that of ORB-SLAM3, with the trajectory tracking accuracy improved by 96.35%. Comparative evaluations demonstrate that the proposed system achieves superior performance in dynamic environments, exhibiting both lower trajectory drift and enhanced positioning precision relative to state-of-the-art dynamic SLAM methods. Full article

Enhancing radar data assimilation at cloud-resolving scales is essential for advancing typhoon analysis and forecasting. This study focuses on Typhoon Linfa, the 10th Pacific Typhoon of 2015, and proposes T-TREC-IS (Typhoon Circulation Tracking Radar Echo by Correlations with Iterative Smoothing), an enhanced version of the T-TREC algorithm. The enhancement incorporates an iterative smoothing constraint into the T-TREC algorithm, which improves the continuity of the retrieved wind field and mitigates the effects of velocity aliasing in radar data, thereby increasing the operational feasibility of the method. Building on this improvement, we evaluate the effectiveness of assimilating the T-TREC-IS-retrieved wind field for analyzing and forecasting Typhoon Linfa. The results demonstrate that the iterative smoothing constraint effectively filters out velocity de-aliasing errors during radar data quality control, enhances wind field intensity near the typhoon core, and retrieves the typhoon circulation more accurately. The refined wind field exhibits improved consistency and continuity, resulting in superior performance in subsequent assimilation analyses and forecasts. Full article

We developed and validated an animation-based assessment (ABA) method for evaluating high school students’ inquiry competencies in Taiwan’s 12-Year Curriculum. Contextualized in atmospheric chemistry involving methane and hydroxyl radicals, ABA integrated dynamic simulations, tiered multiple-choice and open-ended tasks, and process tracking on the CloudClassRoom platform, the assessment focused on measuring two inquiry skills: causal reasoning and critical thinking. The results of 26,823 students revealed that the ABA effectively differentiated student performance across ability levels and academic disciplines, with open-ended items sensitive to higher-order reasoning. Gender difference was not observed, indicating the gender-free design of the developed ABA. While the ABA supports diagnostic insights, limitations need to be addressed, including the underassessment of modeling and creative experimentation skills. Therefore, it is necessary to include open modeling tasks and AI-powered semantic scoring. The developed ABA contributes a scalable, competency-aligned framework for inquiry-based science assessments. Full article