Object Detection in Remote Sensing Images Based on a Scene-Contextual Feature Pyramid Network

Chen, Chaoyue; Gong, Weiguo; Yongliang, Chen; Li, Weihong

doi:10.3390/rs11030339

Cited by 61 publications

(43 citation statements)

References 31 publications

(54 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We adopt the aggregated residual transformations for deep neural networks (ResNeXt-101 32 × 8d) [31] as our backbone of the bottom-up pathway. Due to its superior performance in the field of image processing, it is widely used in many object detectors [4]. The backbone usually has many layers that generate feature maps with the same spatial size and we define these layers as stages.…”

Section: Bottom-up Pathwaymentioning

confidence: 99%

“…With the rapid development of deep convolutional neural networks (CNNs) [1] in recent years, the conventional object detection methods [2,3] have made some remarkable achievements in natural images. However, due to the huge scale variations of the vast majority of objects and the compact distribution of many small objects in remote sensing images, it still remains a tremendous challenge for locating and predicting the target objects [4,5].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Extended Feature Pyramid Network with Adaptive Scale Training Strategy and Anchors for Object Detection in Aerial Images

Guo¹,

Li²,

Gong³

et al. 2020

Remote Sensing

Self Cite

View full text Add to dashboard Cite

Multi-scale object detection is a basic challenge in computer vision. Although many advanced methods based on convolutional neural networks have succeeded in natural images, the progress in aerial images has been relatively slow mainly due to the considerably huge scale variations of objects and many densely distributed small objects. In this paper, considering that the semantic information of the small objects may be weakened or even disappear in the deeper layers of neural network, we propose a new detection framework called Extended Feature Pyramid Network (EFPN) for strengthening the information extraction ability of the neural network. In the EFPN, we first design the multi-branched dilated bottleneck (MBDB) module in the lateral connections to capture much more semantic information. Then, we further devise an attention pathway for better locating the objects. Finally, an augmented bottom-up pathway is conducted for making shallow layer information easier to spread and further improving performance. Moreover, we present an adaptive scale training strategy to enable the network to better recognize multi-scale objects. Meanwhile, we present a novel clustering method to achieve adaptive anchors and make the neural network better learn data features. Experiments on the public aerial datasets indicate that the presented method obtain state-of-the-art performance.Remote Sens. 2020, 12, 784 2 of 22 high-level semantic information at each scale. This structure displays an obvious improvement as a common feature extractor in some practical applications. However, since large-scale objects are usually produced and predicted in the deeper convolution layers of the FPN, the boundaries of these objects might be too fuzzy to obtain accurate regression. Furthermore, the FPN usually predicts small-scale objects in the shallower layers with low semantic information which might not be enough to identify the class of the objects. The designer of the FPN has been aware of this problem and adopted a top-down structure with lateral connections to fuse shallow layers and high-level semantic information to relieve it. However, if the small-scale objects disappear in the deep convolution layers, the context information cues will disappear at the same time.

show abstract

Section: Bottom-up Pathwaymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Extended Feature Pyramid Network with Adaptive Scale Training Strategy and Anchors for Object Detection in Aerial Images

Guo¹,

Li²,

Gong³

et al. 2020

Remote Sensing

Self Cite

View full text Add to dashboard Cite

show abstract

“…The need for such algorithms appears in the efficiency analysis of modern optoelectronic sensors [3]. Similar studies are necessary for the development of methods for tech troubleshooting, appearing in a form of the alternating equipment failures [4]. In modern sections of computational mathematics these methods are required to create algorithms for detecting low-contrast and small-sized objects on aerospace images, and, for example, in signal theory, the same methods are used to estimate the reliability of random fields and point images registration [5 -6].…”

Section: Introductionmentioning

confidence: 99%

Time-optimal algorithms focused on the search for random pulsed-point sources

et al. 2019

View full text Add to dashboard Cite

The article describes methods and algorithms related to the analysis of dynamically changing discrete random fields. Time-optimal strategies for the localization of pulsed-point sources having a random spatial distribution and indicating themselves by generating instant delta pulses at random times are proposed. An optimal strategy is a procedure that has a minimum (statistically) average localization time. The search is performed in accordance with the requirements for localization accuracy and is carried out by a system with one or several receiving devices. Along with the predetermined accuracy of localization of a random pulsed-point source, a significant complicating factor of the formulated problem is that the choice of the optimal search procedure is not limited to one-step algorithms that end at the moment of first pulse generation. Moreover, the article shows that even with relatively low requirements for localization accuracy, the time-optimal procedure consists of several steps, and the transition from one step to another occurs at the time of registration of the next pulse by the receiving system. In this case, the situation is acceptable when during the process of optimal search some of the generated pulses are not fixed by the receiving system. The parameters of the optimal search depending on the number of receiving devices and the required accuracy of localization are calculated and described in the paper.

show abstract

“…In recent years, deep neural networks, especially convolutional neural networks (CNNs), have been widely used in remote sensing areas. They perform incredibly on visual tasks such as scene classification [1,2], change detection [3][4][5], artificial object detection [6] and extraction [7,8]. Among them, extracting buildings, as a set of the most important artificial objects from very high-resolution (VHR) images, is challenging and draws attention from remote sensing communities.…”

Section: Introductionmentioning

confidence: 99%

An Improved Boundary-Aware Perceptual Loss for Building Extraction from VHR Images

Zhang

Gong

et al. 2020

Remote Sensing

Self Cite

View full text Add to dashboard Cite

With the development of deep learning technology, an enormous number of convolutional neural network (CNN) models have been proposed to address the challenging building extraction task from very high-resolution (VHR) remote sensing images. However, searching for better CNN architectures is time-consuming, and the robustness of a new CNN model cannot be guaranteed. In this paper, an improved boundary-aware perceptual (BP) loss is proposed to enhance the building extraction ability of CNN models. The proposed BP loss consists of a loss network and transfer loss functions. The usage of the boundary-aware perceptual loss has two stages. In the training stage, the loss network learns the structural information from circularly transferring between the building mask and the corresponding building boundary. In the refining stage, the learned structural information is embedded into the building extraction models via the transfer loss functions without additional parameters or postprocessing. We verify the effectiveness and efficiency of the proposed BP loss both on the challenging WHU aerial dataset and the INRIA dataset. Substantial performance improvements are observed within two representative CNN architectures: PSPNet and UNet, which are widely used on pixel-wise labelling tasks. With BP loss, UNet with ResNet101 achieves 90.78% and 76.62% on IoU (intersection over union) scores on the WHU aerial dataset and the INRIA dataset, respectively, which are 1.47% and 1.04% higher than those simply trained with the cross-entropy loss function. Additionally, similar improvements (0.64% on the WHU aerial dataset and 1.69% on the INRIA dataset) are also observed on PSPNet, which strongly supports the robustness of the proposed BP loss.

show abstract

Object Detection in Remote Sensing Images Based on a Scene-Contextual Feature Pyramid Network

Cited by 61 publications

References 31 publications

Extended Feature Pyramid Network with Adaptive Scale Training Strategy and Anchors for Object Detection in Aerial Images

Extended Feature Pyramid Network with Adaptive Scale Training Strategy and Anchors for Object Detection in Aerial Images

Time-optimal algorithms focused on the search for random pulsed-point sources

An Improved Boundary-Aware Perceptual Loss for Building Extraction from VHR Images

Contact Info

Product

Resources

About