Cross-modality fusion for pavement scenes is essential for effective automated defect detection in large-scale infrastructure inspections. Current fusion methods often overlook modality coupling relationships, limiting the clear representation of subtle defects and effective suppression of background noise. To address these issues, we propose a novel cross-modality feature decoupling fusion framework specifically designed for pavement scenarios. This framework explicitly models both modality-specific distinctions and inter-modal correlations. Our method leverages frequency-driven decoupling information extracted from depth images to guide the fusion process effectively. For improved feature representation in large-scale automation, we introduce a two-stage training strategy. First, decoupling information is acquired independently; then, it is fused at the pixel level. Additionally, we design an Adaptive Fusion Rate Block (AFRB) to dynamically allocate features during the Feature Allocation Fusion (FAF) stage. By incorporating frequency-driven decoupling, our pixel-level fusion significantly preserves structural clarity and subtle defect details, effectively capturing even non-significant defects. Benchmark experiments confirm that our approach achieves superior visual quality and enhanced defect detection performance. These results underline its effectiveness and potential for automated large-scale pavement inspection systems.

A major obstacle for high-level tasks is the unpredictable image degradation. While several architectures proposed to address this, they fail under extreme degradation. Therefore, we introduce a novel cross-modality pipeline called AIRMamba, designed to holistically and robustly restore images degraded due to extreme weather. Specifically, we devise a strategy that utilizes infrared images to create compact high-frequency priors for the restoration process. Meanwhile, we leverage long-range modeling capability of Mamba to achieve both feature extraction and interaction. We emphasize extracting low-frequency representations from the ground truth and achieving this task through a regression-based approach. Consequently, AIRMamba can achieve reliable restoration through large-gap cross-domain guidance. To facilitate this task, we have constructed a cross-modality restoration benchmark, named WeatherInfrared. Our pipeline is simple, robust, and outperforms several state-of-the-art methods in benchmark evaluations.

The Sea-surface environment is influenced by various factors such as wind, waves, tides, and ocean currents. Interactions among various factors, combined with heterogeneous spatio-temporal distribution, lead to highly uncertain dynamics for sea-surface objects. Accordingly, forecasting sea-surface object trajectories constitutes a highly significant challenge. A common approach is to employ temporal prediction networks to model the trajectories of objects in complex motion. However, these methods do not accurately capture the motion state of an object at any given moment. To address this challenge, we propose a Spatial Representation and Motion Pattern Graph Neural Network (SRMPGNN). We represent the motion states of objects as tokenized textual descriptions and perform spatially aligned mapping and interaction with graph neural features. Moreover, long-horizon sea-surface objects are extremely faint, and we propose Graph Super-Resolution Reconstruction (GSR) to address this challenge. We further design spatial representations and Multi-Frequency Enhancement (ME) to highlight the trajectory characteristics of moving objects on the sea-surface. Through end-to-end training, the SRMPGNN achieves state-of-the-art (SOTA) performance on two benchmarks.

Nearshore multi-object tracking (NMOT) aims to locat and identify nearshore objects. Most approaches accomplish this task using radar and remote-sensing technologies. In contrast, video data can describe the visual appearance of nearshore objects without prior information, such as identity, location, or movement. In this study, we introduce a cross-modality pipeline to address the four major challenges of NMOT. First, we propose introducing a cross-modality bi-attention transformer (CBT) manage the information interaction between RGB and thermal infrared videos effectively. This decoupling and guidance mechanism laid the foundation for our subsequent processes. Next, we integrate the outputs of the backbone with historical frames to extract crucial temporal features. Subsequently, we refine small object detection performance by employing multi-scale feature alignment (MFA). Observations are generated by the transformer decoder. To tackle challenges arising from extensive occlusion and interactions induced by waves in NMOT, we propose guiding modulation (GM), supplemented by low-confidence boxes and multi-point corner momentum (MCM) to facilitate association. Our approach is simple, online, and real-time, showcasing outstanding performance in benchmark evaluations. The open-source implementation of our work is available at https://github.com/Ding-JianGang/Cross-Modality-MOT-in-Nearshore-Environments.

The detection of sealed cracks in pavement images can be complicated by the presence of objects on the pavement that have similar morphology or texture to cracks, leading to false positive results. To address this issue, this paper proposes a High Resolution-Pixel to Pixel (HR-Pix2Pix) pavement distress image enhancement algorithm. The algorithm improves the proportion of effective semantics by reducing the semantic interference caused by pseudo pavement distress such as pavement stains, brake marks, and shadows. The proposed HR-Pix2Pix generative adversarial neural network involves a novel generator architecture for fusing multi-level semantic features. The generator is capable to automatically locate the pseudo-distress in the image and fill (generate) the region as normal pavement texture with reference to the pavement texture surrounding the pseudo-distress. To further improve the quality of the semantic distribution-enhanced images, a hybrid loss function is designed and used to train the network. The results of the study show that by using HR-Pix2Pix enhanced images for sealed crack detection can significantly reduce false detections and improve the accuracy of the detection, consequently.

Pavement life span is often affected by the amount of voids in the base and subgrade soils, especially the soil moisture content. Ground Penetrating Radar (GPR) is one of the desirable techniques to indirectly measure the in-situ soil moisture content through electrical properties of soils. The inversion using transmission line matrix method from GPR data is applied for converting moisture content of the soils. Laboratory and field tests proved satisfactory results.

A ground-wave technique is introduced in this letter to directly extract the pavement permittivity and thickness from measured data of ground-coupled ground-penetrating radar (GPR). Analytic solution, numerical simulation, and experimental test are carried out to validate the method. This technique enables bistatic radar to obtain both thickness and permittivity by just one measurement, which effectively reduces measurement and computation time for GPR applications.