Event cameras generate asynchronous events with extremely high temporal resolution and dynamic range, making them well suited for motion perception in high-speed and challenging environments. Accurate rotational motion estimation is therefore essential for intelligent systems such as autonomous robots and spacecraft navigation platforms. However, existing event-based rotational odometry methods often struggle to balance robustness, estimation accuracy, and real-time performance. In this paper, we propose a novel Time-Aware Mean Minimization (TAMM) method that incorporates the temporal information of events into the construction of motion-corrected spatio-temporal representations. By leveraging a time-aware mean minimization objective, the proposed method exploits both spatial and temporal event information, thereby improving estimation robustness. To further enhance real-time applicability, we introduce a multi-reference time warping strategy that is specifically designed to be compatible with the TAMM optimization framework. This strategy incrementally fuses a small subset of new events with the spatio-temporal image obtained from the previous temporal window, thereby significantly reducing computational cost while maintaining estimation accuracy. Experimental results on both synthetic and real-world datasets demonstrate that the proposed method achieves competitive or improved accuracy in challenging scenarios while substantially improving real-time performance, indicating its effectiveness for real-time event-based rotational motion estimation.
The interaction between tires and soil is widely recognized as a pivotal scientific topic in the field of off-road vehicle engineering, particularly with regard to the examination of tire traction performance when traversing granular wading-terrain. Granular wading-terrain herein refers to two types of granular conditions, which are a stratified terrain with water overlying dry sand, and a mixed water-soil terrain. This study aims to develop a tire-water-particle coupling numerical model based on the finite element method (FEM), smooth particle hydrodynamics (SPH), and discrete element method (DEM) to investigate the traction behavior of an FEM tire model on the SPH-DEM granular wading-terrain. Firstly, the effectiveness of the FEM-DEM, FEM-SPH, and DEM-SPH coupled methods is validated by comparing them with experimental results of the angle of repose, dam break, and water-entry problems. Then, a tire-water-particle FEM-DEM-SPH coupling numerical model is presented and employed to investigate the interaction mechanism between FEM tire models and granular terrain (DEM) with water (SPH), including drawbar pull, tire sinkage, and flow of soil and water particles. Also, the effects of water surface height, granular terrain moisture content, tire inflation pressure, tire load, and slip on the traction behavior of tires on wet terrains are investigated based on the proposed numerical model. The numerical results demonstrate that the FEM-DEM-SPH coupling method can accurately predict the tire-water-particle coupling problem, thereby providing a reliable tool for the design and optimization of tires and off-road vehicles in granular wading-terrain.
Noncooperative maritime targets remain difficult to monitor continuously because AIS, SAR and optical remote sensing are limited by vessel cooperation, revisit time, visibility and sea state. Distributed Acoustic Sensing (DAS) can convert existing submarine optical cables into passive wide area acoustic arrays, but four quadrant heading detection from one dimensional cable records is challenged by strong ocean noise and anisotropic spatiotemporal signatures. Here, we propose a Spatiotemporal Anisotropic Decoupling Detection Transformer (STADETR) for ship heading detection based on DAS. The framework combines MDVMD-Frangi dual branch preprocessing for noise suppression and wake ridge enhancement, a STAD backbone inspired by CSP for separate temporal and spatial feature extraction, and Direction Aware Varifocal Loss (DAVFLoss) to penalize opposite heading errors. On the DAShip Freighter subset, STADETR achieves an mAP50 of 95.91%, exceeding the RT-DETR baseline by 1.35 percentage points after the same preprocessing. Supplementary experiments on four dominant DAShip vessel categories, including Towing vessel, High Speed Craft, Freighter and Tanker, obtain an mAP50 of 94.71%. These results indicate that the proposed balance between accuracy and efficiency is suitable for smuggling prevention and anomalous navigation analysis assisted by DAS, while remaining within the scope of four quadrant heading detection.
Gamification of learning has been regarded as an important pedagogy to promote students’ learning outcomes, and attempts have been made to integrate it into STEM education. However, the influence of gamification of learning on K-12 students’ STEM educational outcomes, such as domain knowledge, higher-order thinking, and affective outcomes, remains under-researched. This study aims to investigate strategies for integrating gamified learning in STEM education, examine its effects on domain knowledge, higher-order thinking, and affective outcomes, and explore K-12 students’ perceived benefits and challenges of implementing gamified learning in STEM education. We systematically reviewed 86 empirical studies involving 7,128 participants from leading bibliographic databases, examined the effects of gamified learning in STEM using a meta-analysis, and investigated the benefits and challenges of implementing gamified learning in STEM education using a qualitative synthesis. Our findings revealed medium to large effect sizes in favour of gamified learning on domain knowledge (Hedges’ g = .823; p < .001), higher-order thinking (Hedges’ g = .880; p < .001), and affective outcomes (Hedges’ g = .577; p < .001) using a random-effects model. The thematic analysis of interview data from 26 articles identified three themes with 15 sub-themes of benefits and three themes with six sub-themes of challenges of implementing gamified learning in STEM education. This review synthesized findings on the impact of gamified learning across STEM learning outcomes and various moderating factors, while also providing rich insights into authentic participant perceptions of its application in STEM education. This study yielded pedagogical and theoretical insights to effectively implement gamified learning in STEM education.
Infrared imaging is indispensable for its ability to penetrate obscurants and visualize thermal signatures, yet its practical use is hindered by the intrinsic limitations of conventional detectors. Nonlinear upconversion, which converts infrared light into the visible band, offers a promising pathway to address these challenges. Here, we demonstrate high-efficiency infrared upconversion imaging using nonlinear silicon metasurfaces. By strategically breaking in-plane symmetry, the metasurface supports a high-Q quasi-bound states in the continuum resonance, leading to strongly enhanced third-harmonic generation (THG) with a conversion efficiency of 3×10−5 at a pump intensity of 10 GW/cm2. Through this THG process, the metasurface enables high-fidelity upconversion of arbitrary infrared images into the visible range, achieving a spatial resolution of ~6 μm as verified using a resolution target and various customized patterns. This work establishes a robust platform for efficient nonlinear conversion and imaging, highlighting the potential of CMOS-compatible silicon metasurfaces for high-performance infrared sensing applications with reduced system complexity.
