Search Results (119)

Restoring images captured in low-light rainy scenes is challenging because brightness degradation and rain corruption are strongly coupled. Enhancing visibility may amplify hidden rain streaks and noise, whereas aggressive deraining can suppress already weak scene structures. Existing cascaded pipelines and general restoration models often struggle to handle this interaction effectively. In this paper, we present the Degradation-Robust Hue Prior Network (DHP-Net), a single-stage framework for low-light rainy image restoration that combines degradation-robust hue prior guidance with perturbation-aware feature modulation. Specifically, DHP-Net extracts multi-scale hue priors to provide stable structural and color cues under coupled degradations, and it injects them into a hierarchical Transformer restoration backbone. To further improve interaction among entangled feature responses, we introduce a Channel-adaptive Attention Perturbation Module that reorganizes intermediate representations before cross-channel aggregation. In this way, the proposed model jointly promotes visibility enhancement, rain removal, and structure preservation within a unified architecture. Extensive experiments on the Low-Light Rain (LLR) benchmark show that DHP-Net achieves 33.14 dB Peak Signal-to-Noise Ratio (PSNR) and 0.9252 Structural Similarity Index Measure (SSIM) on synthetic data and also delivers superior perceptual quality on real-world low-light rainy images, consistently outperforming existing state-of-the-art restoration models. Full article

Image deraining remains a critical challenge in computer vision, as rain streaks significantly degrade visual quality by introducing complex patterns that overlay scene content, impeding applications from autonomous driving to surveillance systems. Existing methods often struggle to balance global scene understanding with precise removal of rain streaks, resulting in either residual rain artifacts or over-smoothed textures. To address this, we propose MPINet, a Multi-Stage Progressive Illumination-Aware Network that integrates illumination awareness and global-context modeling, which is specifically designed for rain-removal tasks. Our architecture features a novel illumination-aware module that generates illumination maps to enhance robustness in the varying lighting conditions commonly encountered during rainy scenarios. The UniMetaFormer core adaptively incorporates global semantic information through dynamic transformations and attention mechanisms, effectively distinguishing between rain streaks and underlying image content. Inspired by MPRNet’s multi-stage restoration framework, our network employs a hierarchical approach with progressive patch-based processing and deep supervision across three stages, enabling efficient integration of our illumination-aware modules while maintaining reasonable model complexity. Experimental results verify the effectiveness of MPINet for rain-removal tasks, demonstrating superior capability in removing rain streaks of varying densities while preserving original image textures. On average, MPINet outperforms MPRNet by approximately 6.5% in terms of PSNR and 1.3% in terms of SSIM across all datasets. Full article

The aim in single-image deraining tasks is to remove rain streaks from degraded images while preserving scene structures and fine details. However, existing deep learning-based methods often face a trade-off between restoration quality and computational efficiency, and many models struggle to capture hierarchical information effectively under complex rain conditions. To address these limitations, we propose a lightweight cross-gated hierarchical transformer for image deraining. The proposed network adopts a five-stage encoder–decoder architecture with Multi-head Feature-wise Attention (MFA) to efficiently model channel-wise dependencies while reducing the computational burden associated with conventional self-attention. In addition, an Enhanced Gated Depthwise Feed-Forward Network (EGDFN) is introduced to obtain refined feature representations with improved efficiency, and a Cross-Level Feature Refinement (CLFR) module is designed to enhance information exchange between corresponding encoder and decoder stages, thereby strengthening hierarchical feature integration and preserving structural details. The network is trained using a single SSIM-based loss, which enhances the structural fidelity of the restored results. Extensive experiments on four synthetic datasets, two real-world datasets, and a downstream semantic segmentation benchmark demonstrate that the proposed method consistently achieves strong restoration performance, producing cleaner outputs with sharper details and improved effectiveness for subsequent vision tasks. Full article

Existing single-image deraining methods often exhibit limited 2D long-range dependency modeling and underexploit frequency-domain priors. To address this, we propose SFQMamba, a dual-branch deraining network based on spatial–frequency feature fusion. The CNN branch employs a Fused Enhance Block (FEB), which integrates multi-scale spatial modeling with global frequency modulation, supported by residual coupling and channel guidance, to suppress rain streaks and recover structural details. Concurrently, the Mamba branch utilizes a Spatial-Aware Selective Fusion Block (SASFB). By incorporating a four-directional scanning mechanism and adaptive path-gating, SASFB extends 1D State Space Models into the 2D domain for content-aware feature fusion. Features from both branches are hierarchically aggregated via concatenation and pointwise convolution. Experiments on the Rain13K and Raindrop datasets show that SFQMamba provides robust restoration. Compared with TransMamba, it obtains improvements of 0.12 dB in PSNR and 0.11% in SSIM, removing dense rain streaks while preserving structural and textural details. Furthermore, on the RainVisDrone benchmark, specifically the medium-rain subset, our method improves YOLOv8s detection by 0.0737 AP, 0.1060 AP${⁠}_{50}$, and 0.0897 AP${⁠}_{75}$ over degraded inputs. These results indicate that the proposed framework benefits both low-level visual restoration and downstream object perception in UAV applications. Full article

This study estimates rainfall disaggregation coefficients for the State of Rio de Janeiro and for the Rio de Janeiro Metropolitan Region (RMRJ) based on automatic rain gauges from the CEMADEN network. A Python-based workflow collected time series, selected stations according to record length, extracted annual extreme events (10 min to 48 h), and calculated sub-daily to daily rainfall ratios for return periods of 2–100 years. The formulations proposed by Pfafstetter and Chen were evaluated through a case study to guide the model selection. In the RMRJ, 109 stations were analyzed and aggregated by municipality, resulting in the metropolitan mean disaggregation coefficient (COERM). The COERM values are close to those proposed by CETESB up to the 30 min–1 h duration range. However, the coefficients were up to 18.8% higher in the duration range between 1 h and 3 h relative to the 24 h rainfall, indicating a stronger temporal concentration of precipitation precisely in durations critical for urban drainage design. Full article

Foreign object detection during unmanned aerial vehicle (UAV) grid inspection suffers from severe visual degradation under adverse weather conditions, such as haze and heavy rain. Existing approaches often struggle to distinguish target textures from weather-induced noise, leading to critical performance drops. We propose the Physics-Prior Complex Wavelet Unrolling Decoupling Module (PCW-UDM) to enable highly robust detection in complex environments. By leveraging the 2D dual-tree complex wavelet transform (2D-DTCWT), our method decouples degraded features into low-frequency and multi-directional high-frequency sub-bands. To tackle haze, we design a Physics-Guided Low-Frequency Dehazing (PGLD) branch that physically inverses the atmospheric scattering process. To combat rain, we introduce the LISTA-Unrolled High-Frequency Deraining (LUHD) branch, which innovatively applies deep unrolled sparse optimisation to remove directional rain streaks without distorting the structural phase. A novel spatio-temporal cross-domain consistency loss further guarantees weather-invariant feature alignment. Extensive evaluations on synthesised adverse datasets and the real-world RTTS dataset prove that our PCW-UDM-equipped network fundamentally overcomes the semantic conflicts of traditional cascaded restoration–detection paradigms, achieving state-of-the-art detection precision and robustness against extreme weather conditions. Full article

In recent years, Transformer architectures have excelled at modeling non-local information. This makes them suitable for image deraining. However, existing methods use dense self-attention. They compute all similarities between query and key tokens. This is inefficient. In practice, this approach can lead to the neglect of the most relevant information and result in a blurring effect of irrelevant representations during the feature aggregation process. To address this issue, this paper proposes an image deraining Transformer based on sparse non-local self-attention. The core of the network consists of multiple non-local feature extraction modules, primarily comprising a sparse self-attention network and a sparse feedforward network along the channel dimension. Specifically, we implement sparse attention by selecting the most useful similarities based on Top-k approximations. Furthermore, we have developed a sparse feedforward network to achieve more accurate representations for high-quality preservation results. Extensive experiments on benchmark datasets have demonstrated the effectiveness of our proposed method. Full article

Image deraining is a crucial preprocessing task for enhancing the robustness of high-level vision systems under adverse weather conditions. However, most of the existing methods are limited to a single RGB color space, and it is difficult to effectively separate high-frequency rain streaks from low-frequency backgrounds, resulting in color distortion and detail loss in the restored image. Therefore, a rain removal network that combines dual-color space and frequency domain priors is proposed. Specifically, the devised network employs a dual-branch Transformer architecture to extract color and structural features from the RGB and YCbCr color spaces, respectively. Meanwhile, a Hybrid Attention Feedforward Block (HAFB) is constructed. HAFB achieves feature enhancement and regional focus through a progressive perception selection mechanism and a multi-scale feature extraction architecture, thereby effectively separating rain streaks from the background. Furthermore, a Wavelet-Gated Cross-Attention module is designed, including a Wavelet-Enhanced Attention Block (WEAB) and a Dual Cross-Attention module (DCA). This design enhances the complementary fusion of structural information and color features through frequency-domain guidance and bidirectional semantic interaction. Finally, experimental results on multiple datasets (i.e., Rain100L, Rain100H, Rain800, Rain12, and SPA-Data) demonstrate that the proposed method outperforms other approaches. Full article

Sensor degradation poses a significant challenge in autonomous driving. During heavy rainfall, interference from raindrops can adversely affect the quality of LiDAR point clouds, resulting in, for instance, inaccurate point measurements. This, in turn, can potentially lead to safety concerns if autonomous driving systems are not weather-aware, i.e., if they are unable to discern such changes. In this study, we release a new, large-scale, multi-modal emulated rain dataset, REHEARSE-3D, to promote research advancements in 3D point cloud de-raining. Distinct from the most relevant competitors, our dataset is unique in several respects. First, it is the largest point-wise annotated dataset (9.2 billion annotated points), and second, it is the only one with high-resolution LiDAR data (LiDAR-256) enriched with 4D RADAR point clouds logged in both daytime and nighttime conditions in a controlled weather environment. Furthermore, REHEARSE-3D involves rain-characteristic information, which is of significant value not only for sensor noise modeling but also for analyzing the impact of weather at the point level. Leveraging REHEARSE-3D, we benchmark raindrop detection and removal in fused LiDAR and 4D RADAR point clouds. Our comprehensive study further evaluates the performance of various statistical and deep learning models, where SalsaNext and 3D-OutDet achieve above $94\%$ IoU for raindrop detection. Full article

The transformer architecture and its attention-based modules have become quite popular recently and are used for solving most computer vision tasks. However, there have been attempts to explore whether other modules can perform equally well with lower computational costs. In this paper, we introduce a nonlinear convolution structure composed of learnable polynomial and Fourier features, which allows better spectral representation with fewer parameters. The solution we propose is in principle feasible for many CNN application fields, and we present its theoretical motivation. Next, to demonstrate the performance of our architecture, and we exploit it for a paradigmatic task: image translation in driving-related scenarios such as deraining, dehazing, dark-to-bright, and night-to-day transformations. We use specific benchmark datasets for each task and standard quality parameters. The results show that our network provides acceptable or better performances when compared to transformer-based architectures, with a major reduction in the network size due to the use of such a nonlinear convolution block. Full article

In supervised training for underwater image enhancement (UIE), many deep learning methods rely on approximate reference images; however, they tend to neglect the inherent uncertainty of these references. To address this problem, this study proposes MFP-PAINet, an underwater image enhancement network, which can integrate multi-dimensional feature fusion with probabilistic adaptive uncertainty modeling. Specifically, histogram equalization, white balance correction, and gamma correction were first implemented to generate preliminary enhanced inputs so as to improve contrast and color balance. Subsequently, a multi-dimensional feature mapping module was used to extract high-dimensional representations and produce multiple feature maps, which were dynamically weighted and adaptively fused by a confidence generation module to initially handle the uncertainty at the feature level. Finally, a probabilistic module was introduced to further model the uncertainty and refine image details. Extensive experiments conducted on eight publicly available datasets demonstrate that the proposed MFP-PAINet outperforms both traditional and deep learning-based UIE methods. For example, on the LSUI dataset, MFP-PAINet achieved SSIM, PSNR, and UIQM scores of 0.854, 23.044, and 3.056, respectively. Furthermore, MFP-PAINet exhibited certain promising performance in image deraining and image dehazing tasks, proving that it can effectively retrieve image details and improve image clarity. Full article

This work introduces Uni-Removal, an innovative two-stage framework that effectively addresses the critical challenge of domain adaptation in unified image restoration. Contemporary approaches often face significant performance degradation when transitioning from synthetic training environments to complex real-world scenarios due to the substantial domain discrepancy. Our proposed solution establishes a comprehensive pipeline that systematically bridges this gap through dual-phase representation learning. In the first stage, we implement a structured multi-teacher knowledge distillation mechanism that enables a unified student architecture to assimilate and integrate specialized expertise from multiple pre-trained degradation-specific networks. This knowledge transfer is rigorously regularized by our novel Instance-Grained Contrastive Learning (IGCL) objective, which explicitly enforces representation consistency across both feature hierarchies and image spaces. The second stage introduces a groundbreaking output distribution calibration methodology that employs Cluster-Grained Contrastive Learning (CGCL) to adversarially align the restored outputs with authentic real-world image characteristics, effectively embedding the student model within the natural image manifold without requiring paired supervision. Comprehensive experimental validation demonstrates Uni-Removal’s superior performance across multiple real-world degradation tasks including dehazing, deraining, and deblurring, where it consistently surpasses existing state-of-the-art methods. The framework’s exceptional generalization capability is further evidenced by its competitive denoising performance on the SIDD benchmark and, more significantly, by delivering a substantial 4.36 mAP improvement in downstream object detection tasks, unequivocally establishing its practical utility as a robust pre-processing component for advanced computer vision systems. Full article

Image restoration tasks such as deraining, deblurring, and dehazing are crucial for enhancing the environmental perception of autonomous vehicles and traffic systems, particularly for tasks like vehicle detection, pedestrian detection and lane line identification. While transformer-based models excel in these tasks, their prohibitive computational complexity hinders real-world deployment on resource-constrained platforms. To bridge this gap, this paper introduces a novel Soft Knowledge Distillation (SKD) framework, designed specifically for creating highly efficient yet powerful image restoration models. Our core innovation is twofold: first, we propose a Multi-dimensional Cross-Net Attention(MCA) mechanism that allows a compact student model to learn comprehensive attention relationships from a large teacher model across both spatial and channel dimensions, capturing fine-grained details essential for high-quality restoration. Second, we pioneer the use of a contrastive learning loss at the reconstruction level, treating the teacher’s outputs as positives and the degraded inputs as negatives, which significantly elevates the student’s reconstruction quality. Extensive experiments demonstrate that our method achieves a superior trade-off between performance and efficiency, notably enhancing downstream tasks like object detection. The primary contributions of this work lie in delivering a practical and compelling solution for real-time perceptual enhancement in autonomous systems, pushing the boundaries of efficient model design. Full article

High-quality image restoration (IR) is a fundamental task in computer vision, aiming to recover a clear image from its degraded version. Prevailing methods typically employ a static inference pipeline, neglecting the spatial variability of image content and degradation, which makes it difficult for them to adaptively handle complex and diverse restoration scenarios. To address this issue, we propose a novel adaptive image restoration framework named Hybrid Convolutional Transformer with Dynamic Prompting (HCTDP). Our approach introduces two key architectural innovations: a Spatially Aware Dynamic Prompt Head Attention (SADPHA) module, which performs fine-grained local restoration by generating spatially variant prompts through real-time analysis of image content and a Gated Skip-Connection (GSC) module that refines multi-scale feature flow using efficient channel attention. To guide the network in generating more visually plausible results, the framework is optimized with a hybrid objective function that combines a pixel-wise L1 loss and a feature-level perceptual loss. Extensive experiments on multiple public benchmarks, including image deraining, dehazing, and denoising, demonstrate that our proposed HCTDP exhibits superior performance in both quantitative and qualitative evaluations, validating the effectiveness of the adaptive restoration framework while utilizing fewer parameters than key competitors. Full article

Acoustic emission (AE) is a passive non-destructive testing (NDT) method that allows for online monitoring of oil tank bottom corrosion without production shutdown. However, AE signals are susceptible to ambient noise interference, causing the AE inspection system to mistakenly identify noise as corrosion signals, which significantly reduces AE inspection performance. Therefore, it is important to distinguish between AE signals caused by corrosion and those caused by noise. To address this, an AE inspection platform for vertical atmospheric tank corrosion is established. Six common noise sources in field AE inspections, including mechanical vibration and friction, fluid and raining disturbance, external impacts, and oil leakage are simulated. The impacts of these noises on AE location events are analyzed. Variational mode decomposition (VMD) and dispersion entropy (DE) are used to extract multi-domain features of AE signals. An improved distance evaluation (IDE) algorithm is then introduced to obtain a highly correlated feature subset. A support vector machine (SVM) model optimized by the bald eagle search (BES) algorithm is proposed to identify different noise sources. Field experiments demonstrate that for mechanical friction, external impacts, and effective corrosion signals, the proposed method achieves identification accuracy of 92.95% and 94.00% in the training and test sets, respectively. This proves the reliability of the BES-SVM model, which uses multi-domain features for AE source identification in oil tank bottom corrosion inspections. Moreover, the impacts of the optimization algorithm, feature selection algorithm, and feature type on AE source identification are further investigated. Full article