Multispectral object detection is a crucial technology in remote sensing image processing, particularly in low-light environments. Most current methods extract features at a single scale, resulting in the fusion of invalid features and the failure to detect small objects. To address these issues, we propose a multispectral object detection network based on multilevel feature fusion and dual feature modulation (GMD-YOLO). Firstly, a novel dual-channel CSPDarknet53 network is used to extract deep features from visible-infrared images. This network incorporates a Ghost module, which generates additional feature maps through a series of linear operations, achieving a balance between accuracy and speed. Secondly, the multilevel feature fusion (MLF) module is designed to utilize cross-modal information through the construction of hierarchical residual connections. This approach strengthens the complementarity between different modalities, allowing the network to improve multiscale representation capabilities at a more refined granularity level. Finally, a dual feature modulation (DFM) decoupling head is introduced to enhance small object detection. This decoupled head effectively meets the distinct requirements of classification and localization tasks. GMD-YOLO is validated on three public visible-infrared datasets: DroneVehicle, KAIST, and LLVIP. DroneVehicle and LLVIP achieved mAP@0.5 of 78.0% and 98.0%, outperforming baseline methods by 3.6% and 4.4%, respectively. KAIST exhibited an MR of 7.73% with an FPS of 61.7. Experimental results demonstrated that our method surpasses existing advanced methods and exhibits strong robustness.

Lane detection is a vital component of intelligent driving systems, offering indispensable functionality to keep the vehicle within its designated lane, thereby reducing the risk of lane departure. However, the complexity of the traffic environment, coupled with the rapid movement of vehicles, creates many challenges for detection tasks. Current lane detection methods suffer from issues such as low feature extraction capability, poor real-time detection, and inadequate robustness. Addressing these issues, this paper proposes a lane detection algorithm that combines an online re-parameterization ResNet with a hybrid attention mechanism. Firstly, we replaced standard convolution with online re-parameterization convolution, simplifying the convolutional operations during the inference phase and subsequently reducing the detection time. In an effort to enhance the performance of the model, a hybrid attention module is incorporated to enhance the ability to focus on elongated targets. Finally, a row anchor lane detection method is introduced to analyze the existence and location of lane lines row by row in the image and output the predicted lane positions. The experimental outcomes illustrate that the model achieves F1 scores of 96.84% and 75.60% on the publicly available TuSimple and CULane lane datasets, respectively. Moreover, the inference speed reaches a notable 304 frames per second (FPS). The overall performance outperforms other detection models and fulfills the requirements of real-time responsiveness and robustness for lane detection tasks.

Molybdenum silicides are attractive high-temperature structural materials because of their excellent thermal stability and outstanding oxidation resistance at high temperatures. First-principles calculations were employed to investigate the effect of alloying elements (Cr, Nb, V, W, Al, Ga, and Ge) on the mechanical properties of Mo₃Si. The structural stabilities of doped Mo₃Si were calculated, showing that the Pm-3n structure was stable at the investigated low-doping concentration. The calculated elastic constants have also evaluated some essential mechanical properties of doped Mo₃Si. Cr- and V-doping decreased the elastic modulus, while Al- and Nb-doping slightly increased the shear and Young’s modulus of Mo₃Si. Furthermore, V-, Al- and Nb-doping decreased the B/G and Poisson ratio, suggesting that these elements could form strong covalent bonds, and decrease shear deformation and alloy ductility. Based on the three-dimensional contours and two-dimensional projection of the elastic modulus, Cr- and V-doping exhibited a significant influence on the anisotropy of the shear and Young’s modulus. According to charge density and density of states, the electronic structures of alloyed Mo₃Si were further analyzed to reveal the doping effects.

In order to tackle the issues of poor mobility and unstable traction of electric vehicles on sandy landscapes, this research develops a high-accuracy numerical model for wheel–sand interaction relying on the Discrete Element Method (DEM). An innovative parameter calibration procedure is proposed herein, which optimizes the sand contact parameters. This reduces the error between the simulated and measured angles of repose to merely 1.2% and substantially improves the model’s reliability. The model was then used to systematically compare the performance of a 205/55 R16 slick tire with a treaded tire on sand. Simulations demonstrate that at a 30% slip ratio, the treaded tire exhibited significantly higher traction and greater sinkage than the slick tire. This indicates that tread patterns enhance traction mechanically by increasing the contact area and promoting shear deformation of the sand. The trends of traction with slip ratio and the corresponding sand flow patterns showed excellent agreement with experimental observations, which validated the simulation approach. This research provides an efficient and accurate tool for evaluating tire-sand interaction, providing critical support for the design and control of electric vehicles on complex terrains.