Global Context Parallel Attention for Anchor-Free Instance Segmentation in Remote Sensing Images

Liu, Xinyu; Di, Xiaoguang

doi:10.1109/lgrs.2020.3023124

Cited by 10 publications

(6 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Currently, attention mechanism is widely used in instance segmentation tasks, such as Chen et al [31] uses the boundary attention mechanism to optimize the feature extraction effect. Liu et al [32] constructed a global context parallel attention mechanism for instance segmentation. Shang et al [33] established an instance-level context attention mechanism to obtain the correlation between different instance objects.…”

Section: B Visual Attention Mechanism For Instance Segmentationmentioning

confidence: 99%

RPFNet: Recurrent Pyramid Frequency Feature Fusion Network for Instance Segmentation in Side-Scan Sonar Images

Wang

Zhang

et al. 2024

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Side-scan sonar (SSS) is an essential acoustic sensor device for obtaining underwater information. The instance segmentation of sonar images can effectively locate and detect underwater objects. Although various CNN-based frameworks have achieved promising results in natural image instance segmentation, the noise interference, highlight shadow, and blurred edge in sonar images bring more significant challenges for sonar image instance segmentation. To solve these problems, we propose a novel recurrent pyramid frequency feature fusion network (RPFNet), which mainly consists of the pyramid frequency feature fusion network (PFN), recurrent residual attention mechanism (RRAM), mask prediction module, and semantic segmentation module. By enhancing and fusing different frequency features of SSS images, PFN can effectively extract fine-grained features and reduce background information interference. The RRAM uses residual structure and attention mechanism to enhance the correlation of different frequency features and improve nonlinear feature representation ability. The mask prediction and semantic segmentation modules are used to discriminate object instance categories and generate corresponding semantic segmentation results. We extensively evaluate the proposed framework on the real-scenes sonar image dataset. The evaluation results demonstrate that the proposed method outperforms the state-ofthe-art methods in various evaluation metrics. Code is available at https://github.com/darkseid-arch/SonarRPFNet.

show abstract

Section: B Visual Attention Mechanism For Instance Segmentationmentioning

confidence: 99%

RPFNet: Recurrent Pyramid Frequency Feature Fusion Network for Instance Segmentation in Side-Scan Sonar Images

Wang

Zhang

et al. 2024

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Despite the flourishing development of instance segmentation in the nature scene, there are only few studies [29]- [35] in RSIs. Feng et al [29] introduced the sequence local context module to address the confusion between densely arranged ships.…”

Section: A Instance Segmentationmentioning

confidence: 99%

“…HQ-ISNet [34] introduces the HR-FPN to maintain highresolution feature maps in the network and designs a tiny network to refine the original mask branch. Liu et al [35] embedded a global context parallel attention module into the anchor-free instance segmentation framework to capture the global information. Different from the methods [29]- [32] that only focused on a single category (e.g.…”

Section: A Instance Segmentationmentioning

confidence: 99%

“…In the past two decades, with the development of convolutional neural networks (CNN) [12], [13], many instance segmentation architectures [14]- [28] have been proposed and achieved outstanding performance in the natural scene. However, few researchers [29]- [35] focus on the instance segmentation in RSIs and the available methods just apply the instance segmentation models designed for the natural images to the RSIs, without taking into account the characteristics of RSIs such as the complex background and diversely scaled instances. Specifically, RSIs typically contain a highly complex background area that may interfere with the region of interest.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Semantic Attention and Scale Complementary Network for Instance Segmentation in Remote Sensing Images

Zhang

Zhu

et al. 2021

Preprint

View full text Add to dashboard Cite

In this paper, we focus on the challenging multicategory instance segmentation problem in remote sensing images (RSIs), which aims at predicting the categories of all instances and localizing them with pixel-level masks. Although many landmark frameworks have demonstrated promising performance in instance segmentation, the complexity in the background and scale variability instances still remain challenging for instance segmentation of RSIs. To address the above problems, we propose an end-to-end multi-category instance segmentation model, namely Semantic Attention and Scale Complementary Network, which mainly consists of a Semantic Attention (SEA) module and a Scale Complementary Mask Branch (SCMB). The SEA module contains a simple fully convolutional semantic segmentation branch with extra supervision to strengthen the activation of interest instances on the feature map and reduce the background noise's interference. To handle the undersegmentation of geospatial instances with large varying scales, we design the SCMB that extends the original single mask branch to trident mask branches and introduces complementary mask supervision at different scales to sufficiently leverage the multi-scale information. We conduct comprehensive experiments to evaluate the effectiveness of our proposed method on the iSAID dataset and the NWPU Instance Segmentation dataset and achieve promising performance.

show abstract

“…The anchor-free methods are also introduced in remote sensing images processing. Liu and Di [20] embedded a global context parallel attention module into the anchor-free instance segmentation framework to capture the global information. Polar Template Mask [21] used a nonuniform angle polar coordinate system to solve the problem with masks not being smooth and accurate enough due to the uniform angle sampling of the ship targets with large aspect to achieve competitive accuracy.…”

Section: Introductionmentioning

confidence: 99%

Anchor-Free SAR Ship Instance Segmentation With Centroid-Distance Based Loss

Gao¹,

Huo²,

Wang³

et al. 2021

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Instance segmentation methods for synthetic aperture radar (SAR) ship imaging have certain unsolved problems:(1) Most of the anchor-based detection algorithms encounter difficulties in tuning the anchor-related parameters and high computational costs. (2) Different tasks share the same features without considering the differences between tasks, leading to mismatching of the shared features and inconsistent training targets. (3) Common loss functions for instance segmentation cannot effectively distinguish the positional relationships between ships with the same degree of overlap. In order to alleviate these problems, we first adopt a lightweight feature extractor and an anchor-free convolutional network, which effectively help to reduce computational consumption and model complexity. Secondly, to fully disseminate feature information, a dynamic encoder-decoder is proposed to dynamically transform the shared features to task-specific features in channel and spatial dimensions. Thirdly, a novel loss function based on centroid distance is designed to make full use of the geometrical shape and positional relationship between SAR ship targets. In order to better extract features from SAR images in complex scenes, we further propose the dilated convolution enhancement module (DEM), which utilizes multiple receptive fields to take full advantage of the shallow feature information. Experiments conducted on the SAR Ship Detection Dataset (SSDD) prove that the method proposed in this paper is superior to the other state-of-the-art algorithms in terms of instance segmentation accuracy and model complexity.

show abstract

Global Context Parallel Attention for Anchor-Free Instance Segmentation in Remote Sensing Images

Cited by 10 publications

References 28 publications

RPFNet: Recurrent Pyramid Frequency Feature Fusion Network for Instance Segmentation in Side-Scan Sonar Images

RPFNet: Recurrent Pyramid Frequency Feature Fusion Network for Instance Segmentation in Side-Scan Sonar Images

Semantic Attention and Scale Complementary Network for Instance Segmentation in Remote Sensing Images

Anchor-Free SAR Ship Instance Segmentation With Centroid-Distance Based Loss

Contact Info

Product

Resources

About