Model-Based Deep Network for Single Image Deraining

Li, Pengyue; Tian, Jinshou; Tang, Yandong; Wang, Guolin; Wu, Chengdong

doi:10.1109/access.2020.2965545

Cited by 9 publications

(3 citation statements)

References 42 publications

(54 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since vehicles may also drive in extreme meteorological environments such as the rain or haze, data augmentation by adding rain and fog in images will improve the network performance under the extreme weather. Here, we introduce a nonlinear raindrop model [36], as shown below…”

Section: Data Augmentationmentioning

confidence: 99%

Traffic Intersection Re-Identification Using Monocular Camera Sensors

Liu

Deng

Huang

et al. 2020

Sensors

View full text Add to dashboard Cite

Perception of road structures especially the traffic intersections by visual sensors is an essential task for automated driving. However, compared with intersection detection or visual place recognition, intersection re-identification (intersection re-ID) strongly affects driving behavior decisions with given routes, yet has long been neglected by researchers. This paper strives to explore intersection re-ID by a monocular camera sensor. We propose a Hybrid Double-Level re-identification approach which exploits two branches of Deep Convolutional Neural Network to accomplish multi-task including classification of intersection and its fine attributes, and global localization in topological maps. Furthermore, we propose a mixed loss training for the network to learn the similarity of two intersection images. As no public datasets are available for the intersection re-ID task, based on the work of RobotCar, we propose a new dataset with carefully-labeled intersection attributes, which is called “RobotCar Intersection” and covers more than 30,000 images of eight intersections in different seasons and day time. Additionally, we provide another dataset, called “Campus Intersection” consisting of panoramic images of eight intersections in a university campus to verify our updating strategy of topology map. Experimental results demonstrate that our proposed approach can achieve promising results in re-ID of both coarse road intersections and its global pose, and is well suited for updating and completion of topological maps.

show abstract

Section: Data Augmentationmentioning

confidence: 99%

Traffic Intersection Re-Identification Using Monocular Camera Sensors

Liu

Deng

Huang

et al. 2020

Sensors

View full text Add to dashboard Cite

show abstract

“…It contains hand-crafted regularizers that mention the preceding knowledge of a solution, like repeatability, high-frequency probability of the rain streaks and the image"s piecewise smoothness [14]. Nevertheless, such model-based mechanisms are unsuitable for rainy conditions since the degradation process can become more complicated [15]. To solve this issue, learning-based schemes are utilized to learn the essential attributes from the data like convolutional filters, stochastic distributions and Gaussian mixture models (GMMs) [16][17][18].…”

Section: Introductionmentioning

confidence: 99%

Rain Streaks Removal in Images using Extended Generative Adversarial-based Deraining Framework

Gogulamudi¹,

Mahalakshmi²,

Sreeram³

2023

IJACSA

View full text Add to dashboard Cite

The visual quality of photographs and videos can be negatively impacted by various weather conditions, such as snow, haze, or rain, affecting the quality of the images and videos. Such impacts may greatly affect outdoor vision systems that rely on image/video data. It has recently drawn a lot of interest to remove rain streaks from a single image. Several deep learning-based methods have been introduced to address the issue of removing rain streaks from a single image. Still, the efficiency of rain streak removal with enhanced quality is challenging. Hence, a novel deep-learning method is introduced for rain streak removal. The proposed Extended Generative Adversarial based De-raining (Ex_GADerain) is the enhanced version of a traditional Generative adversarial network (GAN). The proposed Ex_GADerain introduced a Self-Attention based Convolutional Capsule Bidirectional Network (SA-CCapBiNet) based generator for enhancing the rain streaks removal process. Also, the loss function estimation using the adversarial loss and the mean absolute error loss minimizes the information loss during training. The minimal information loss enhances the generalization capability of Ex_GADerain, and hence the enhanced performance is acquired. The quality assessment of a derained image based on various assessment measures like SSIM, PSNR, RMSE, and DSSIM improved performance compared to the conventional rain streak removal methods. The maximal SSIM and PSNR acquired by the Ex_GADerain are 0.9923 and 26.7052, respectively. The minimal RMSE and DSSIM acquired by the Ex_GADerain are 0.9367 and 0.0051, respectively.

show abstract

“…This method is more time-consuming for heavy rain streaks. Li et al (2020) designed a novel channel attention U-dense network for rain pixel detection and a residual dense block for rain streak removal. The U-dense network is used to achieve pixelwise estimation accuracy and rain features are exploited by a dense block.…”

mentioning

confidence: 99%

Performance analysis of a dual stage deep rain streak removal convolution neural network module with a modified deep residual dense network

Jayaraman,

Chinnusamy

2022

View full text Add to dashboard Cite

The visual appearance of outdoor captured images is affected by various weather conditions, such as rain patterns, haze, fog and snow. The rain pattern creates more degradation in the visual quality of the image due to its physical structure compared with other weather conditions. Also, the rain pattern affects both foreground and background image information. The removal of rain patterns from a single image is a critical process, and more attention is given to remove the structural rain pattern from real-time rain images. In this paper, we analyze the single image deraining problem and present a solution using the dual stage deep rain streak removal convolutional neural network. The proposed single image deraining framework primarily consists of three main blocks: a derain streaks removal CNN (derain SRCNN), a modified residual dense block (MRDB), and a six-stage scale feature aggregation module (3SFAM). The ablation study is conducted to evaluate the performance of various modules available in the proposed deraining network. The robustness of the proposed deraining network is evaluated over the popular synthetic and real-time data sets using four performance metrics such as the peak signal-to-noise ratio (PSNR), the feature similarity index (FSIM), the structural similarity index measure (SSIM), and the universal image quality index (UIQI). The experimental results show that the proposed framework outperforms both synthetic and real-time images compared with other state-of-the-art single image deraining approaches. In addition, the proposed network takes less running and training time.

show abstract

Model-Based Deep Network for Single Image Deraining

Cited by 9 publications

References 42 publications

Traffic Intersection Re-Identification Using Monocular Camera Sensors

Traffic Intersection Re-Identification Using Monocular Camera Sensors

Rain Streaks Removal in Images using Extended Generative Adversarial-based Deraining Framework

Performance analysis of a dual stage deep rain streak removal convolution neural network module with a modified deep residual dense network

Contact Info

Product

Resources

About