Search

Recently, object detection in natural images has made a breakthrough, but it is still challenging in oriented ship detection for remote sensing imagery. Considering some limitations in this task, such as uncertain ship orientation, unspecific features for locating and classification in the complex optical environment, and multiplicative speckle interference of synthetic aperture radar (SAR), we propose an oriented ship detector based on the pairwise branch detection head and adaptive SAR feature enhancement. The details are as follows: (1) Firstly, the ships with arbitrary directions are described with a rotated ground truth, and an oriented region proposal network (ORPN) is designed to study the transformation from the horizontal region of interest to the rotated region of interest. The ORPN effectively improved the quality of the candidate area while only introducing a few parameters. (2) In view of the existing algorithms that tend to perform classification and regression prediction on the same output feature, this paper proposes a pairwise detection head (PBH) to design parallel branches to decouple classification and locating tasks, so that each branch can learn more task-specific features. (3) Inspired by the ratio-of-average detector in traditional SAR image processing, the SAR edge enhancement (SEE) module is proposed, which adaptively enhances edge pixels, and the threshold of the edge is learned by the channel-shared adaptive thresholds block. Experiments were carried out on both optical and SAR datasets. In the optical dataset, PBH combined with ORPN improved recall by 5.03%, and in the SAR dataset, the overall method achieved a maximum F1 score improvement of 6.07%; these results imply the validity of our method. [ABSTRACT FROM AUTHOR]

Recently, deep-learning methods have yielded rapid progress for object detection in synthetic aperture radar (SAR) imagery. It is still a great challenge to detect ships in SAR imagery due to ships' small size and confusable detail feature. This article proposes a novel anchor-free detection method composed of two modules to deal with these problems. First, for the lack of detailed information on small ships, we suggest an adaptive feature-encoding module (AFE), which gradually fuses deep semantic features into shallow layers and realizes the adaptive learning of the spatial fusion weights. Thus, it can effectively enhance the external semantics and improve the representation ability of small targets. Next, for the foreground–background imbalance, the Gaussian-guided detection head (GDH) is introduced according to the idea of soft sampling and exploits Gaussian prior to assigning different weights to the detected bounding boxes at different locations in the training optimization. Moreover, the proposed Gauss-ness can down-weight the predicted scores of bounding boxes far from the object center. Finally, the effect of the detector composed of the two modules is verified on the two SAR ship datasets. The results demonstrate that our method can effectively improve the detection performance of small ships in datasets. [ABSTRACT FROM AUTHOR]

TE is a critical and difficult problem that tries to map traffic with various requirements to paths in dynamic communication networks. The emerging SDN enables centralized TE optimization with a global view. From the architectural viewpoint, ICN facilitates TE from many aspects, such as in-network caching which can reduce redundant traffic and content-awareness which can extract prior knowledge of content type directly. Thus, we leverage ICN to optimize SDN TE. However, ICN brings more complexities and dynamics to the network environment, which makes model-based TE methods inefficient. Inspired by recent advances in applying artificial intelligence techniques to solve complex online control problems, we investigate deep learning for content-awareness and DRL for TE decision. In addition, we propose a parallel online learning mechanism to safely utilize DRL which has trial-and-error nature. Results show that our proposal significantly improves network performance in terms of total network throughput, bandwidth utilization, and load balance. [ABSTRACT FROM AUTHOR]

• We present the Multi-scale Recurrent Attention Gated Fusion Network (MRAGFN) as an end-to-end solution for tackling the image dehazing task. • We develop the Global Feature Extraction Module (GFEM), which utilizes dilated convolutions of different receptive fields to effectively extract complex features from the input image. • We introduce the Dual Attention Fusion (DAF) module to effectively implement the attention mechanism and create haze-relevant feature maps. • We design the Feature Enhancing Module (FEM), drawing inspiration from the SOS boosting algorithm. • We propose the Recurrent Attention Gated Fusion (RAGF) module, which incorporates attention and gating mechanisms. This module progressively refines features by giving weight to relevant information while reducing redundancy. The purpose of single image dehazing is to eliminate the bad influence of haze on images, so as to maintain more scene information of images. In recent years, the convolutional neural networks (CNN) have made significant contributions to single image dehazing. However, the visual quality of dehazed images still needs to be further improved. In view of the problems of single-scale shallow image feature extraction and the insufficient use of intermediate layer features in existing dehazing networks, we propose an end-to-end Multi-scale Recurrent Attention Gated Fusion Network (MRAGFN) to address the image dehazing task. We cascade three Dual Attention Fusion (DAF) modules to progressively form three haze-relevant features map, meanwhile, we adopt downsampling operation on the input to produce global feature map, which are used to weight the three feature maps to compensate for the missing of single-scale feature information. We present Feature Enhancement Module (FEM) to enhance the feature representation ability of these weighted feature maps. We design Recurrent Attention Gated Fusion (RAGF) module by adding attention mechanism and gating mechanism to gradually obtain more refined features based on these weighted features while eliminating redundant features. Experimental results on different hazy images demonstrate that the proposed dehazing network can restore the haze-free images and perform better than the state-of-the-art dehazing networks in terms of the objective indicators (such as PSNR, SSIM) and the subjective visual quality. [ABSTRACT FROM AUTHOR]

In the past two decades, traditional hand-crafted feature based methods and deep feature based methods have successively played the most important role in image classification. In some cases, hand-crafted features still provide better performance than deep features. This paper proposes an innovative network based on deep learning integrated with binary coding and Sinkhorn distance (DBSNet) for remote sensing and texture image classification. The statistical texture features of the image extracted by uniform local binary pattern (ULBP) are introduced as a supplement for deep features extracted by ResNet-50 to enhance the discriminability of features. After the feature fusion, both diversity and redundancy of the features have increased, thus we propose the Sinkhorn loss where an entropy regularization term plays a key role in removing redundant information and training the model quickly and efficiently. Image classification experiments are performed on two texture datasets and five remote sensing datasets. The results show that the statistical texture features of the image extracted by ULBP complement the deep features, and the new Sinkhorn loss performs better than the commonly used softmax loss. The performance of the proposed algorithm DBSNet ranks in the top three on the remote sensing datasets compared with other state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

• The SSEResNet feature fusion backbone network based on ResNet101 is designed by using SENet attention module and CSPNet enhancement module. • A visual monitoring method of laser cleaning quality based on improved Cascade R-CNN is proposed by using SSEResNet, BiFPN and Soft-NMS. • A visual monitoring method of surface quality after laser cleaning based on SSEResNet101 regression model is proposed by reforming SSEResnet backbone network. • The effects of four different learning rate decay strategies on the two proposed deep learning models are studied. In order to protect the substrate during the cleaning process as well as evaluate the cleaning effect and surface quality after laser cleaning of aircraft coatings, a visual monitoring method based on deep learning is proposed. In this paper, the data sets of "flame recognition-cleaning quality evaluation" and "optical image-surface roughness" are constructed and data enhancement is performed. The SSEResNet backbone network which can effectively extract the details of the input image is designed by using the feature fusion method. The Cascade R-CNN object detection model is improved by using SSEResNet, BiFPN and Soft-NMS, and the SSEResNet101 regression model which can directly predict surface roughness from optical images is proposed based on ResNet101. Model comparison and ablation experiments show that the above two deep learning models proposed by us have excellent detection ability and regression prediction performance, and can realize flame recognition, cleaning effect judgment during laser cleaning as well as post-cleaning surface quality evaluation. In this paper, the effects of four different learning rate decay strategies on the models are further studied. The results show that the training effect of CosineAnnealing with warm restart method is the best. In SSEResNet101 model, the training mean square error (MSE) loss is 0.0249, the mean absolute error (MAE) is 0.278μm, and the test MAE is 0.245μm; In improved Cascade R-CNN model, the mean average precision (mAP) value of intersection over union (IoU=0.6) reaches 93.6%. [ABSTRACT FROM AUTHOR]