Search Results (72)

Autonomous driving systems rely on a sophisticated pipeline of artificial intelligence models to perceive, predict, and plan in dynamic environments. This review presents a systematic analysis of the machine learning and deep learning models underpinning vehicle autonomy, spanning classical convolutional neural networks (CNNs) for object detection and semantic segmentation to recurrent and Transformer-based architectures for trajectory prediction and motion planning. It also provides a critical examination of the autonomous vehicle sensor stack, including cameras, LiDAR, radar, ultrasonics, and GNSS/IMU as data acquisition systems, highlighting modality-specific AI challenges such as monocular depth estimation, 3D point cloud processing, and radar Doppler interpretation. The evolution of perception and decision-making pipelines is reviewed, contrasting modular architectures with end-to-end learning paradigms that directly map raw sensor data to control commands, and discussing their trade-offs in interpretability, safety assurance, and robustness to rare edge cases. We further survey specialized hardware accelerators and heterogeneous automotive SoCs designed to meet stringent real-time and power constraints. Industrial strategies are compared, including multi-modal sensor fusion and vision-centric approaches based on large-scale imitation learning. Finally, we identify open challenges related to robustness under adverse conditions, domain shift, causal ambiguity, and the need for interpretable and certifiable AI in safety-critical autonomous driving systems. Full article

Autonomous Haulage Systems (AHS) have significantly transformed surface mining operations by improving safety, productivity, and operational consistency. Currently, AHS predominantly rely on vehicle-centric perception architectures. Onboard LiDAR, radar, cameras, and Global Navigation Satellite Systems (GNSS) perform sensing, interpretation, and decision-making within individual systems. These processes enable collision avoidance and path tracking. However, they are limited in their ability to consider the broader, dynamic mining environment characterized by dust, terrain degradation, geotechnical instability, heterogeneous traffic, and rapidly evolving operational conditions. This paper presents a systematic review of dynamic vision systems of AHS in surface mining. It critically analyzes the transition from autonomy to interconnected, ecosystem-aware intelligence. The review synthesizes literature from mining automation, robotics, intelligent transportation systems, and multi-agent perception. It assesses sensing technologies, perception algorithms, sensor fusion strategies, and environmental robustness techniques. Attention is focused on the limitations of egocentric perception models in complex surface mining ecosystems. Building on identified gaps, the paper proposes a conceptual framework for Ecosystem-Centric Dynamic Vision (ECDV). Perception is enhanced through integration with fleet communication networks, dispatch systems, digital twins, geotechnical monitoring platforms, and environmental sensing infrastructure. The framework outlines a multi-layer architecture enabling cooperative perception, predictive hazard modeling, and risk-aware decision support at the mine-wide level. The review concludes by outlining a research agenda to transition from vehicle autonomy to ecosystem intelligence in surface mining. It highlights opportunities in cooperative perception, adaptive sensor fusion under degraded visibility, and digital-twin-integrated predictive safety systems. Full article

The Vortex Ring State (VRS) poses a catastrophic aerodynamic threat to coaxial dual-rotor unmanned aerial vehicles (UAVs). Traditional reactive detection mechanisms provide insufficient altitude for recovery, while existing data-driven diagnostics are severely bottlenecked by data leakage, extreme class imbalance, and a lack of physical interpretability. To bridge these gaps, this paper proposes a physics-informed multimodal deep learning framework that transitions from post-occurrence detection to proactive early warning. We establish a 1.5 s precursor window—creating a three-class ordinal state space—to provide the flight control system with critical intervention time for differential rotor recovery. We developed a novel ViT-LSTM architecture (MTSF-Net) to fuse continuous seven-channel onboard-recorded data (comprising three-axis acceleration, three-axis angular velocity, and barometric vertical velocity), which are subsequently transformed into Continuous Wavelet Transform (CWT) spectrograms. To ensure real-time unidirectional inference while preserving absolute physical vibration scales across heterogeneous sensors, a Calibrated Benchmark Normalization (CBN) strategy is introduced. Furthermore, a Hybrid Ordinal Loss is proposed to mitigate the extreme sample imbalance (<0.5%) of the precursor state by penalizing asymmetric aerodynamic degradation. Evaluated under a strict sortie-based isolation protocol, the proposed system achieves an exceptional test accuracy of 98.26% and an unprecedented precursor recall of 100%. Notably, it completely eliminates fatal missed detections (VRS predicted as Normal) and false-positive VRS predictions triggered by precursor states. Finally, Gradient-weighted Class Activation Mapping (Grad-CAM) is utilized to verify that the multimodal sensor processing pipeline successfully anchors onto authentic physical vibration frequencies rather than artifactual noise, laying a rigorous, interpretable foundation for intelligent aviation safety systems. Full article

The detection technology of crop diseases and pests is transitioning from single sensor monitoring to intelligent perception and multimodal fusion. This paper follows the PRISMA 2020 standard and systematically reviews the relevant core literature. This paper systematically summarizes the development history of spectral sensing technology and analyzes the physical mechanisms of hyperspectral and multispectral imaging in early identification of crop diseases. The focus is on the architectural evolution of deep learning models, including lightweight convolutional neural networks (CNNs), vision transformers (ViTs) with long-range dependency modeling capabilities, and the efficient computing state space model Mamba. In addition, the research progress of spatial spectral joint learning, heterogeneous data fusion, and vision-language models (VLMs) in improving system robustness and interpretability are introduced. By synthesizing the integrated applications of UAV remote sensing, Internet of Things (IoT) edge computing and intelligent robots in staple and cash crops, this paper summarizes the implementation of the integrated system of perception, decision-making and execution. To address the issues of insufficient cross-domain generalization ability and uneven allocation of computing resources in existing models, this paper provides perspectives on the future development of agricultural artificial intelligence (AI) towards foundation model-driven, edge-intelligent collaboration, and green sustainable direction, which can provide theoretical reference for engineering applications in the field of intelligent plant protection. Full article

The field of pose estimation is a major problem in computer vision, enabling the direct transformation of an input image into a hierarchical representation of the human skeleton for application in the fields of virtual/augmented reality and human–machine interaction tasks. Research in this field has exploded between 2018 and 2025, with traditional taxonomies such as 2D versus 3D or top-down versus bottom-up no longer sufficient to capture the essence of the evolution of ideas. To solve this problem, we propose a conceptual review in the field of pose estimation, focusing on the intellectual evolution of methods and architecture rather than the standard flat classifications of papers. We divide recent advances into five structural pillars: Representation, which traces the evolution from pixel coordinate regression to heatmaps and probabilistic representation; Architecture, which analyzes the transition from multi-stage CNNs to transformers and state space models (SSMs); Ambiguity and Generalization, which analyzes how self-supervised, uncertainty-aware, and diffusion models address 3D depth ambiguity, occlusion, and domain gaps by modeling multiple plausible poses and reducing dependence on fully supervised in-the-wild 3D labels; Context Extension, which covers temporal dynamics, multi-view fusion, and potential sensors; and Applications, which links algorithms to efficiency, privacy, and foundation models. By providing an in-depth detailing of these pillars, we provide a unified view of the evolution of research paradigms that define human pose estimation and enable the identification of future problems and solutions in pose estimation and human-centered tasks. Full article

Achieving long-term, continuous, and accurate localization for Unmanned Aerial Vehicles (UAVs) in outdoor GNSS-denied environments where pre-existing reference maps are available is challenging. To this end, this paper proposes a Cross-view Particle Filter Localization (CPFL) framework. Unlike existing particle filter approaches that rely on inertial sensors for state propagation or sparse semantic labels for observation updates, CPFL is a vision-driven solution. This framework introduces specific adaptations into the two core stages of particle filtering: In the motion propagation stage, it achieves visual state transition by calculating a feature-based inter-frame homography mapping to estimate the 2D global relative motion components, eliminating the dependency on inertial priors; in the observation correction stage, a Dual-Granularity Adaptive Gating (DGAG) cross-view network is designed to mitigate perceptual aliasing and generate discriminative absolute position weights for the particles. By fusing these two stages through a filter mechanism, the framework transforms unbounded cumulative drift into bounded absolute localization errors. Furthermore, addressing the measurement deficiencies of traditional single-frame metrics, this paper also proposes a Trajectory Continuity Index (TCI@d) tailored for continuous localization tasks. Experiments on the real-world MAFS dataset confirm that this framework achieves a mean localization error of 5.28 m and a localization success rate of 89.7% under a 10-m threshold. Compared with mainstream vision-only algorithms and IMU-fusion baselines, this framework demonstrates lower mean errors and improved trajectory continuity, validating its effectiveness for long-term robustness. Full article

The rapid development of autonomous vehicles is based mainly on their ability to accurately perceive their environment, where artificial intelligence and computer vision act as the core of environmental perception. In this regard, deep learning-based perception architectures have revolutionized the field of autonomous driving. However, as the use of single sensors fails to ensure reliability in complex scenarios, multimodal sensor fusion has become an essential part of modern deep learning architectures. In this context, covering the literature from 2020 to 2025, we analyze the transition from traditional Convolutional Neural Networks (CNNs) to modern Vision Transformers (ViTs) and explore data fusion design methodologies at various processing levels. In addition, significant limitations related to adverse weather conditions and dynamic environments, computational resources and overall quality and management of data are identified. The conducted comparative analysis indicates that vision-transformer and multimodal fusion methodologies provide higher accuracy in perception tasks but at the cost of increased computational requirements and sensor synchronization challenges. Finally, it becomes clear that achieving full autonomy requires further research in subjects such as collaborative perception, unsupervised domain adaptation and the creation of lightweight models, thus offering a roadmap for future developments. Full article

The automated inspection of aluminum profile surface defects, which heavily relies on data acquired by machine vision sensors, is a critical task in industrial quality control. Addressing the current challenges of intense background texture interference and the difficulty in detecting defects with extreme aspect ratios on aluminum profiles, this research puts forward a complete end-to-end defect detection algorithm named WMC-DFINE (WIFA-MKSS-CSFF-DFINE) based on the DFINE framework. First, a Wavelet-Integrated Frequency Attention (WIFA) module is introduced, which utilizes a discrete wavelet transform to decouple features into the frequency domain, thereby dynamically suppressing high-frequency background noise and enhancing defect edge responses. Second, a Cross-Scale Feature Fusion (CSFF) module based on dual-channel pooling is designed to ensure the continuity of defect features, thereby resolving the semantic misalignment issue in traditional fusion. Third, a Multi-Kernel Strip Shuffle (MKSS) module is incorporated, utilizing decomposed convolution kernels to capture the geometric features of slender scratches. Finally, a knowledge distillation strategy is employed to transfer structured knowledge from a complex teacher model to a lightweight student model. Experiments on the Tianchi aluminum defect dataset demonstrate that WMC-DFINE achieves a mAP of 82.1%, which surpasses algorithms including YOLOv12, RT-DETR, and the baseline model DFINE. Furthermore, the distilled student model, WMC-DFINE-distill, improves the mAP by 3.2% compared to DFINE, reduces parameter count by 47%, and achieves an inference speed of 59.75 FPS on the experimental equipment. The proposed method effectively resolves the problem of balancing background suppression and defect detail feature preservation, offering a practical and efficient scheme for real-time industrial defect inspection. Full article

To address the multi-sensor collaborative calibration challenges in laser 3D projection systems, a pose calibration method integrating binocular vision and laser ranging is proposed. A multi-coordinate system fusion framework encompassing the camera coordinate system, galvanometer coordinate system, and workpiece coordinate system is established. Through the calculation of reference pose matrices and real-time transformations, adaptive calibration under arbitrary workpiece placements is achieved. Experimental results demonstrate that within a working range of 1.5–2.5 m, the calibration error is 45.5 $\mathsf{\mu }$m, meeting the high-precision requirements of aerospace precision machining and assembly. Full article

Background: Myocardial ischemia (MI) is a major cause of morbidity and mortality worldwide and requires timely and reliable detection. Although Spin-Exchange Relaxation-Free (SERF) magnetocardiography (MCG) provides femtotesla-level sensitivity for identifying non-linear cardiac repolarization anomalies, its clinical deployment is currently impeded by the computational bottlenecks inherent to portable edge platforms. Methods: We propose a “Sensor-to-Image” Edge-AI framework that links quantum sensing with computer vision. Single-channel SERF-MCG signals from a large cohort of 2118 subjects (1135 Healthy, 983 Ischemia) were transformed into phase-space images using three distinct encoding modalities: Recurrence Plots (RP), Gramian Angular Summation Fields (GASF), and Markov Transition Fields (MTF). These visual representations were subsequently analyzed by a streamlined MobileNetV3-Small architecture, optimized for low-latency inference. To maximize diagnostic precision, an adaptive weighted fusion mechanism was engineered to combine the chaotic specificity captured by RP with the morphological sensitivity of GASF through a validation-optimized fixed global weighting strategy. Results: In our experiments, the fusion model achieved an Area Under the Curve (AUC) of 0.865, which was higher than the 1D-CNN baseline (AUC 0.857) and the single-modality models. Notably, the fusion strategy significantly elevated sensitivity to 88.3% while maintaining a specificity of 66.5%. Although specificity is moderate, this trade-off prioritizes high sensitivity to minimize false negatives in pre-hospital screening scenarios. The average inference time was 4.7 ms per sample on a standard CPU, suggesting suitability for real-time Point-of-Care (PoC) scenarios under further on-device validation. Conclusions: The results suggest that multi-view phase-space fusion can capture subtle spatio-temporal changes associated with ischemia. The proposed lightweight framework may support the development of portable SERF-MCG systems with embedded AI screening. Full article

Visual quality assessment is entering a new frontier as media evolve from static images to temporally dynamic videos and 3D content. These visual signals are typically captured by sensing devices such as cameras and depth sensors, whose acquisition characteristics significantly influence perceptual quality. Traditional quality models, including distortion-centric and regression-based approaches, perform well on conventional degradations but struggle to evaluate higher-level attributes such as semantic plausibility and structural coherence in modern AI-generated and multimodal scenarios. The emergence of large multimodal models (LMMs), including vision–language models (VLMs) and multimodal large language models (MLLMs), reshapes the evaluation paradigm by enabling semantic grounding, instruction-driven assessment, and explainable reasoning. This survey presents a unified perspective on visual quality assessment for sensor-captured visual data across image, video, and 3D modalities. We review conventional deep learning approaches and recent LMM-based methods, highlighting how multimodal fusion and language-conditioned reasoning transform quality assessment from scalar prediction to perceptual intelligence. Finally, we discuss key challenges and future opportunities for building efficient, robust, and sensor-aware visual quality assessment systems. Full article

Surface cracking is a key visible indicator of deterioration in concrete infrastructure and is routinely captured by vision sensors during field inspections. To translate inspection imagery into actionable maintenance information, crack delineation must be accurate at the pixel level and robust to challenging conditions where cracks are slender, discontinuous, low-contrast, and easily confused with joints, stains, texture patterns, and illumination artifacts. This study proposes a lightweight CNN–Mamba hybrid segmentation framework built upon Vm-unet for reliable crack mapping under heterogeneous inspection scenarios and resource-constrained deployment. The framework couples boundary-sensitive convolutional features with long-range state-space representations via a spatially modulated convolution design, refines skip-connection features using reciprocal co-modulation attention to suppress background interference, and enhances cross-scale interactions through a decoder interaction fusion scheme to preserve fine-crack continuity and sharp boundaries. Experiments on a multi-source composite dataset and public benchmarks show consistent improvements over representative CNN-, Transformer-, and Mamba-based baselines. The proposed method achieves 80.11% mIoU and 82.05% Dice on the composite dataset, while maintaining an efficient accuracy–cost trade-off (36.049 GFLOPs, 25.991 M parameters). The resulting crack masks provide a dependable basis for inspection-driven quantitative assessment and maintenance decision support. Full article

Accurate meteorological visibility estimation is critical to the safety and reliability of transportation and environmental monitoring systems. Despite the prevalence of deep learning, models often struggle with the non-linear visual degradation caused by varying atmospheric conditions and a scarcity of instrument-calibrated datasets. This study makes two primary contributions. First, we introduce the Hong Kong Chu Hai College Visibility Dataset (HKCHC-VD) comprising 11,148 high-resolution images paired with precise visibility measurements from a Biral SWS-100 sensor. Second, we propose a Range-Aware Attention Framework (RAT-Attn), an adaptive attention mechanism that translates classical range-specific atmospheric modeling into differentiable deep learning operations. This is a domain-specific architectural optimization that integrates a dual-backbone architecture (CNN and Vision Transformer) with a learnable threshold mechanism. This design enables the model to dynamically prioritize spatial and channel-wise features based on estimated visibility intervals, specifically targeting the non-linear visual degradation unique to fog and haze. Experimental results demonstrate that our proposed approach outperforms existing baselines, including VisNet and landmark ANN-based methods. The ResNet + ViT (spatial-threshold) variant achieves the most balanced performance, recording a Mean Squared Error (MSE) of 5.87 km², a Mean Absolute Error (MAE) of 1.65 km, and a classification accuracy of 87.07%. In critical low-visibility conditions (0 to 10 km), the framework reduces regression error by over 75% compared to the baselines. These results confirm that range-aware adaptive feature fusion is essential for robust meteorological estimation in real-world environments. Full article

Unmanned Aerial Vehicle imagery often suffers from severe spatial detail degradation due to sensor limitations and motion blur, hindering downstream vision tasks. To address this, we propose a lightweight super-resolution framework leveraging a Tiny-Transformer backbone enhanced by a multi-domain feature fusion strategy. Specifically, we jointly model spatial structural semantics and frequency domain texture priors via a cross-domain fusion attention mechanism, enabling coordinated restoration of global consistency and local details. Extensive experiments demonstrate that our method outperforms state-of-the-art approaches on standard benchmarks, achieving significant gains in Peak Signal-to-Noise Ratio and structural similarity index while maintaining low computational cost. Notably, the model exhibits superior robustness in reconstructing high-frequency textures common in aerial scenes. This work provides an efficient, deployable solution for enhancing visual fidelity in resource-constrained applications such as urban planning and precision agriculture. Full article

Automotive radar has emerged as a pivotal technology for navigation in GNSS-denied environments, offering superior robustness to adverse weather and fluctuating lighting conditions compared to vision or LiDAR-based sensors. Despite these advantages, the inherent sparsity and noise of radar measurements often lead to degraded estimation accuracy and system reliability. To address these challenges, various radar-based localization frameworks have been explored, ranging from optimization-based and Extended Kalman Filter (EKF) approaches fused with Inertial Measurement Units (IMUs) to point cloud registration techniques like Iterative Closest Point (ICP). While filter-based methods are favored in multi-sensor fusion for their proven stability, ICP is widely utilized for high-precision pose estimation in point-cloud-centric systems. In this study, we propose a novel Radar-Inertial Odometry (RIO) framework that synergistically integrates ICP-based relative pose estimation with model-based sensor fusion. The proposed methodology leverages relative transformations derived from ICP alongside ego-velocity estimations obtained from radar Doppler measurements. To effectively incorporate relative ICP constraints, a stochastic cloning technique is implemented to augment previous states and their associated covariances, ensuring that the uncertainty of historical poses is explicitly accounted for. The performance of the proposed method is validated using public open-source datasets, demonstrating higher localization accuracy and more consistent performance compared to existing algorithms used for comparison. Full article