Search Results (3,856)

3D multi-object tracking (MOT) for autonomous driving remains challenging due to frequent identity switches in crowded scenes, trajectory fragmentation during occlusions, and the difficulty of adapting association strategies to varying scene complexities. While existing methods rely on fixed geometric or appearance-based associations, they struggle to handle ambiguous cases and detection failures. We present an adaptive multi-level 3D MOT framework that achieves robust tracking through three key innovations: (1) multi-granularity temporal modeling that captures both fine-grained short-term motion and coarse long-term trends via dual-scale spatio-temporal attention, enabling accurate motion prediction across different object dynamics; (2) Transformer-based Appearance Association that employs cross-attention to model global inter-object relationships, resolving ambiguous associations in crowded scenarios where geometric cues alone fail; and (3) scene-adaptive learned thresholds that automatically adjust association strictness based on object density, motion complexity, and occlusion levels, avoiding the one-size-fits-all limitations of fixed thresholds. Our hierarchical four-level tracking strategy progressively handles cases from easy geometric matching (Level 1) to complex interval-frame recovery (Level 4), with SOT-based virtual detection generation bridging detector failures. Extensive experiments on the nuScenes benchmark demonstrate state-of-the-art performance. Full article

With the rapid advancement of autonomous driving in modern society, dense pedestrian detection technology has encountered performance bottlenecks. To address this, we propose a robust and lightweight pedestrian detection algorithm, RMP-YOLO, designed to efficiently detect small, occluded, and low-light objects. Firstly, RFAConv is utilized as the core component of the backbone network, combining standard convolution with attention mechanisms and using group convolution to extract features from the spatial receptive field. Secondly, MobileViTv3 is introduced into the backbone to combine CNNs with Transformers. The model is further enhanced by adjusting feature fusion, introducing residual connections, and optimizing local representation with deep convolutional layers. Finally, the PIoUv2 loss function is employed for bounding-box regression, significantly reducing detection errors for small-scale pedestrians in crowded environments. Experimental results demonstrate that RMP-YOLO improves mAP@0.5 by 1.3% on a custom dataset and 0.91% on the WiderPerson dataset. Crucially, it maintains high efficiency with only 3.71 million parameters and 6.29 GFLOPs, meeting the deployment requirements for low computational power and high precision. Full article

The planning of trajectories for Connected Autonomous Vehicles (CAVs) represents a pivotal aspect of autonomous driving technologies, enabling secure navigation within traffic environments. Traditional models for trajectory control primarily focus on the efficiency and safety of individual vehicles but often overlook the dynamics involved in vehicle-to-vehicle and vehicle-to-infrastructure interactions. This study introduces a novel concept, the “driving risk field,” which imposes constraints on vehicular movement within designated road spaces to enhance safety. A vehicle dynamics model is developed, employing a non-linear fifth-degree polynomial to approximate the trajectory curves, with optimization performed using the Sequential Quadratic Programming (SQP) method. The efficacy of the optimized model is validated through simulations on the Prescan/Simulink plat Full article

The inherent redundancy in vehicle sensor data, coupled with constrained onboard resources and stringent latency requirements, renders traditional bit-oriented transmission paradigms inefficient for autonomous-driving perception tasks. Semantic communication offers a promising direction by shifting the focus from bit-level fidelity to task-level information delivery. In this paper, we propose a unified framework that integrates hierarchical transmission and online scheduling for Internet of Vehicles (IoV)-oriented collaborative perception. The proposed hierarchy separates information into two complementary layers: a coarse metadata layer (object bounding boxes) for latency-critical awareness, and fine-grained visual semantics (multi-scale region-of-interest (ROI) patches) for perception-intensive tasks. We formulate an online scheduling problem that jointly exploits Age of Information (AoI) and Channel State Information (CSI) to dynamically decide what to transmit and at what fidelity under per-frame budget constraints. To address cross-scheme fairness, we report resource utilization under a fixed kbps/fps physical budget and evaluate robustness using a combination of a lightweight task-proxy metric and COCO-style Average Recall (AR100) under ROI-only evaluation. The hierarchical transmission architecture, combined with AoI awareness, reduces global semantic staleness by approximately 78%. The Lyapunov-based online scheduler enables intelligent, signal-to-noise ratio (SNR)-adaptive switching between coarse and fine semantic levels, ensuring robust perception under varying channel quality. Under strict physical-budget constraints and unreliable channel conditions, joint source-channel coding (JSCC) exhibits significantly stronger task robustness than conventional schemes: at 0 dB SNR, the task-proxy detection rate improves by nearly 47 percentage points over the uncoded baseline. Full article

The Controller Area Network (CAN) remains the backbone of in-vehicle communication, but its lack of built-in security exposes safety-critical systems to cyberattacks. This paper presents a real-time, reconfigurable, specification-driven intrusion detection system (IDS) implemented on a custom test bench that emulates an EV powertrain. The CAN traffic captured from the four-ECU setup formed the dataset used in this study. The IDS enforces a compact, reconfigurable ruleset covering timing bounds, jitter envelopes, identifier whitelists, frame format, data length code (DLC) compliance, bus-load thresholds, application-level CRC, and alive-counter verification. The IDS achieves detection times below 2 ms with false positive rates under 1% for injection, denial of service (DoS), and fuzzy attacks, even at CAN bus loads up to 70%, while microcontroller resource usage remains within the constraints of automotive-grade devices, supporting deployment in embedded environments. The main contributions of this study are as follows: (i) a validated and reproducible EV powertrain test bench with millisecond-level timing, (ii) a deployable and easily reconfigurable ruleset with deterministic runtime, and (iii) a latency-oriented evaluation framework that is portable across automotive microcontroller platforms. The EV powertrain dataset v1.0 was released in a public GitHub repository to facilitate reproducible research and enable future benchmarking studies. Full article

Pulse-width-modulated (PWM) automotive headlights enhance nighttime event-based camera detection, yet systematic parameter optimization for vulnerable road user detection remains unexplored. This study evaluates PWM frequency, duty cycle, light distribution, ego-vehicle speed, and ambient lighting under European New Car Assessment Programme-inspired crossing scenarios for cyclist and pedestrian detection. Results establish performance ranging from substantial improvements to severe degradation relative to continuous illumination. Cyclist detection achieves robust performance with high-frequency modulation across light distributions, while low-frequency operation with low beam produces severe degradation through background noise accumulation. Pedestrian detection requires high beam with street lighting enabled; low beam universally fails regardless of modulation parameters. Limited parameter combinations achieve simultaneous improvements for both targets. Detection performs optimally on retroreflective surfaces, while low-reflectivity clothing limits capability, requiring target-specific optimization. Full article

To reduce the conservatism of local trajectory planning in dynamic road scenarios caused by redundant projection of predicted trajectories, this paper proposes a hierarchical hybrid trajectory-planning framework with a time-to-arrival (TTA)-driven dynamic risk-filtering mechanism. In the Frenet coordinate system, road boundaries, ego states, and static and dynamic obstacles are represented uniformly to construct an S–L fused risk field and an S–T spatiotemporal interaction graph, enabling the filtering of temporally irrelevant conflict regions based on TTA relationships. At the path-planning layer, risk-guided adaptive sampling is integrated with dynamic programming and quadratic programming to improve search efficiency and trajectory quality. At the speed-planning layer, spatiotemporal coordination is achieved through non-uniform discretization, safe-corridor extraction, and speed-profile optimization. Simulation results show that the proposed method generates safe, smooth, continuous, and executable local trajectories in scenarios involving static-obstacle avoidance, adjacent-vehicle cut-ins, non-motorized road-user crossings, and mixed multi-obstacle interactions, while reducing unnecessary deceleration and detours. Ablation results further indicate that adaptive sampling reduces the number of DP search nodes by approximately 50% and the average planning time by about 30%, while maintaining a nearly unchanged minimum safety distance. These findings demonstrate that the proposed framework effectively suppresses redundant conflict regions and improves planning efficiency, solution feasibility, and motion continuity without compromising safety. Full article

Rapid urbanization has led to a significant increase in land use carbon emission (LCE), putting great pressure on ecological security. The coupling relationship between LCE and the ecological security index (ESI) is the key to sustainable development. Based on land use/cover change (LUCC) and Open-Data Inventory for Anthropogenic Carbon dioxide (ODIAC) data, the LCE of the Jilin Border Cities (JLBCs) from 2013 to 2023 was estimated. Twenty-seven indicators were selected from both natural and socioeconomic aspects to evaluate the ESI using the Driving forces–Pressure–State–Impact–Response–Management (DPSIRM) model. The spatial interaction between LCE and ESI was analyzed using the coupling degree model and spatial autocorrelation. The results show that from 2013 to 2023, the main LCE areas in the JLBCs were concentrated in central urban districts, while the total LCE remained negative but exhibited a clear upward trend. The ESIs in Tonghua City and Baishan City have continued to improve, but those in Yanbian Autonomous Prefecture have gradually deteriorated, with ecological security warnings intensifying progressively toward the east. The spatial variation in the LCE–ESI coupling degree is significant, predominantly exhibiting low coupling with differences across scales. Within the study area, coupling degree shows a strong positive correlation, revealing distinct spatial clustering patterns dominated by low clusters and cold spots. Future efforts should focus on promoting low-carbon development models, strengthening protection and restoration, while implementing targeted measures to enhance the overall ecology of JLBCs. Full article

In sandstorm environments, a large number of suspended particles in the air absorb and scatter light, causing strong color bias, low contrast, and blurred details in images. These degradations reduce the reliability of computer vision applications in surveillance systems, intelligent transportation systems, unmanned aerial vehicle monitoring, and outdoor autonomous driving systems. A complete sandstorm image enhancement method was developed in this study by combining sky detection, color correction, contrast enhancement, and adaptive dark channel prior (ADCP) dehazing. The Lab color space was used to correct the color bias. The L channel was enhanced using normalized gamma correction and contrast-limited adaptive histogram equalization to improve brightness and contrast. Then, the sky region is detected to avoid over-processing, preserving the natural appearance of the sky region. Finally, ADCP is applied to non-sky regions for further dehazing. Experiments show that the proposed method provides better subjective and objective performance compared to other algorithms. Full article

Image semantic segmentation is essential in autonomous driving, medical imaging, and remote sensing. While convolutional neural networks (CNNs) excel at local feature extraction and spatial structure modeling, their limited receptive fields restrict the capture of long-range dependencies and global semantic consistency. Transformers provide strong global modeling through self-attention but often lack local inductive bias and show weaker generalization on small datasets. To address these limitations, this paper proposes a Multi-Scale Context-aware Network (MSC-Net) for image semantic segmentation. Under an encoder–decoder framework, MSC-Net combines a convolutional backbone with a Multi-Scale Self-Attention module to integrate the complementary strengths of CNNs and attention mechanisms. The backbone extracts local texture and structural information and can adopt architectures such as MobileNet, Xception, DRN, and ResNet, while the attention module captures long-range dependencies and multi-scale contextual information. This design improves cross-layer feature collaboration, multi-scale feature fusion, and boundary quality while maintaining computational efficiency. Experimental results show that MSC-Net achieves 38.8% mIoU and 98.4% ACC under comparable computational settings. Compared with SegFormer and DeepLabV3+, the model improves mIoU by approximately +3.0 and +3.3 percentage points, respectively, while reducing FLOPs and parameter size. Full article

Video question answering remains a challenging task that requires a sophisticated understanding of both visual content and temporal dynamics across video sequences. Current approaches typically rely on fixed temporal processing strategies and uniform frame-selection mechanisms, which fail to adapt to the diverse requirements of different question types and may overlook critical visual information. We propose AdaSeViLA, an adaptive framework that enhances video understanding through two key innovations: Adaptive Temporal Window Selection (ATWS) that dynamically adjusts the number of processed frames (3–12 frames) based on question-type classification, and Object-importance-Aware Frame Selection (OAFS) that combines global relevance with local visual saliency for enhanced frame identification. Our approach intelligently allocates computational resources based on question complexity while maintaining high accuracy through improved frame-selection mechanisms. Extensive experiments on three challenging VideoQA benchmarks demonstrate that AdaSeViLA achieves superior performance: 87.4% accuracy on MM-AU (+2.7% over SeViLA), 73.6% on NExT-QA (+0.4% improvement), and 61.6% on STAR (+0.6% gain), while providing up to 4× computational speedup for short-term tasks. These results validate the effectiveness of adaptive temporal processing and object-aware selection in advancing video question answering capabilities. Full article

Pedestrian trajectory prediction is a fundamental task in autonomous driving and mobile robotics, where accurate forecasting requires modeling of both social interactions and scene-related constraints. However, existing methods typically rely on a fixed interaction modeling strategy, which may be insufficient under heterogeneous crowd densities. To address this limitation, we propose DAGH-Net, a density-adaptive gated hybrid network for pedestrian trajectory prediction. Built upon an SR-LSTM (State Refinement for LSTM) backbone, the proposed framework integrates two complementary reasoning pathways: a data-driven social interaction branch and a hybrid knowledge graph branch that encodes structured relational priors among pedestrians, obstacles, and walkable regions. A local-density-conditioned gating mechanism is further introduced to adaptively fuse these features according to the surrounding crowd condition of each pedestrian. This design helps suppress redundant interaction cues in sparse settings while strengthening socially compliant and scene-consistent reasoning in dense or conflict-prone environments. Experimental results on the ETH (Eidgenössische Technische Hochschule Zürich) and UCY (University of Cyprus) benchmarks, evaluated using Mean Average Displacement (MAD) and Final Average Displacement (FAD), show that DAGH-Net improves the average MAD and FAD by 1.6% and 4.2%, respectively, compared with SR-LSTM. Ablation studies further support the complementary contributions of the hybrid knowledge graph and the density-adaptive gating mechanism. We also discuss the limitations of the current density formulation and benchmark scale, which suggest several directions for future improvement. Full article

Complex weather degrades both perception reliability and tire–road adhesion, thereby reducing the safety margin and responsiveness of intelligent driving longitudinal control. This study proposes a reproducible evaluation method for adverse weather operational design domains based on parameter perturbation testing and comprehensive assessment. Snow, fog, and rain are graded using standard quantitative thresholds and are coupled with road slipperiness to construct a weather–road state set. A mechanism-oriented indicator system, a combined subjective–objective weighting strategy, and a multi-level fuzzy comprehensive evaluation model are then used to generate quantitative capability scores. The method is validated on a co-simulation framework integrating vehicle–sensor simulation, a driving simulator, and a digital-twin testing environment using representative autonomous emergency braking scenarios. Results show that increasing weather severity, decreasing road adhesion, and higher initial speed reduce the post-braking safety margin and prolong collision-response time. The proposed method differentiates performance across weather–road states and provides quantitative support for test-coverage planning and capability boundary calibration in adverse weather operational design domains. Full article

To meet the high-fidelity testing environment requirements for autonomous driving system development, driving simulators are gradually evolving from tools that “only provide scenes and interaction interfaces” into integrated verification platforms for autonomous driving capabilities. These simulators, in particular, need to feature testable and scalable decision-making modules. However, the autonomous driving functions in existing driving simulators mostly rely on rule-based or simplified model approaches, which are inadequate for depicting the complex interactions in real-world traffic and fail to meet the personalized decision-making needs under various driving styles. To address these challenges, this paper designs and implements a trajectory data-driven personalized autonomous driving decision system, using drone aerial imagery as the core data source to provide realistic background traffic flow and human-like decision-making capabilities. The proposed system can be interpreted as an integrated decision–planning–control framework deployed within a high-fidelity driving simulation platform. It consists of a driving style classification module based on drone trajectory data, a personalized decision module integrating inverse reinforcement learning and dynamic game theory, and a planning and control module. First, a natural driving database is built using 4997 real vehicle trajectories, and prior features of different driving styles are extracted through trajectory feature engineering and an improved K-means++ method. Based on this, a personalized decision-making framework that combines dynamic game theory and maximum entropy inverse reinforcement learning is proposed, aiming to learn the preference weights of different driving styles in terms of safety, comfort, and efficiency. Furthermore, the Dueling Network Architecture (DuDQN) is used to generate human-like lane-changing strategies. Subsequently, a real-time closed-loop execution of personalized decisions in the simulation platform is achieved through fifth-order polynomial trajectory planning, lateral Linear Quadratic Regulator (LQR) control, and longitudinal cascade Proportional–Integral–Derivative (PID) control. Experimental results show that the personalized decision model trained with drone data can realistically reproduce vehicle decision-making behaviors in natural traffic flows within the simulation environment and generate autonomous driving strategies that are highly consistent with different driving styles. This significantly enhances the humanization and personalization capabilities of the autonomous driving module in the driving simulator. Full article

To address the bottlenecks of low sample efficiency and poor control accuracy in traditional single-layer reinforcement learning during autonomous driving overtaking, this paper proposes an overtaking decision and control system based on hierarchical reinforcement learning to decouple complex tasks in spatial and temporal dimensions. A heterogeneous two-layer architecture is constructed, where the upper layer adopts the Proximal Policy Optimization algorithm to generate macroscopic discrete decisions, while the lower layer employs Twin Delayed Deep Deterministic Policy Gradient combined with Long Short-Term Memory to achieve smooth continuous control of steering and acceleration by perceiving temporal features of dynamic obstacles. A composite reward mechanism, integrating hard safety constraints and soft efficiency incentives, is designed to balance safety, efficiency, and comfort. Experimental results in complex scenarios with multiple interfering vehicles and random lane-changing behaviors demonstrate that the proposed system improves the training convergence speed by approximately 30% within 500,000 steps compared to single-layer algorithms. In tests across varying traffic densities, the system achieves a 98.3% success rate in medium-density scenarios with a collision rate of only 0.6%. In high-density challenges, the success rate remains above 95%, with the collision rate reduced by about 80% compared to baseline models. Furthermore, the lateral control deviation is strictly limited to within 0.2 m, and the longitudinal safety distance remains stable above 5 m. This system provides a robust, high-efficiency paradigm for autonomous overtaking. Full article