Multistage Attention ResU-Net for Semantic Segmentation of Fine-Resolution Remote Sensing Images

Li, Rui; Zheng, Shunyi; Duan, Chenxi; Su, Jianlin; Zhang, Ce

doi:10.1109/lgrs.2021.3063381

Cited by 91 publications

(48 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To further confirm the effectiveness of the proposed MANet, we compare our method with state-of-the-art approaches presented in the literature. Specifically, the comparative methods not only include the scaling attention mechanism i.e., SE module [2] and CBAM [4] but also consider the simplified dot-product attention mechanism i.e., EAM [3], FAM [11], and LAM [15]. Besides, several comparative networks are also taken into comparison, including the DANet [5] which utilizes the conventional dot-product attention mechanism and other receptive-field-enlarging, i.e., PSPNet [12], DeepLabV3+ [14], as well as EaNet [17].…”

Section: Quantitative Comparison Diverse Methodsmentioning

confidence: 99%

“…As both space and time consumption of the standard dotproduct attention mechanism increase quadratically with the input size, several research has devoted to simplify the complexity of the attention mechanism, including the efficient attention mechanism (EAM) [3], the fast attention mechanism (FAM) [11], and the linear attention mechanism (LAM) [15]. As shown in Table Ⅵ, the proposed KAM achieves the best accuracy compared with other simplified dot-product attention mechanism, due to the appropriate simplified scheme adopted.…”

Section: ) Comparison With Simplified Dot-product Attentionmentioning

confidence: 99%

“…In the field of remote sensing, DL methods have been introduced and implemented for land cover and land use classification [32]. Compared with vegetation indices, which are based on physical and mathematical concepts and hand-coded from spectral bands only, DL methods can mine different kinds of information including temporal periods, spectra, spatial context and the interactions among different land cover categories [15].…”

mentioning

confidence: 99%

See 2 more Smart Citations

Multiattention Network for Semantic Segmentation of Fine-Resolution Remote Sensing Images

Zheng

Zhang

et al. 2022

IEEE Trans. Geosci. Remote Sensing

Self Cite

126

View full text Add to dashboard Cite

Semantic segmentation of remote sensing images plays an important role in a wide range of applications including land resource management, biosphere monitoring and urban planning. Although the accuracy of semantic segmentation in remote sensing images has been increased significantly by deep convolutional neural networks, several limitations exist in standard models. First, for encoder-decoder architectures such as U-Net, the utilization of multi-scale features causes the underuse of information, where low-level features and high-level features are concatenated directly without any refinement. Second, longrange dependencies of feature maps are insufficiently explored, resulting in sub-optimal feature representations associated with each semantic class. Third, even though the dot-product attention mechanism has been introduced and utilized in semantic segmentation to model long-range dependencies, the large time and space demands of attention impede the actual usage of attention in application scenarios with large-scale input. This paper proposed a Multi-Attention-Network (MANet) to address these issues by extracting contextual dependencies through multiple efficient attention modules. A novel attention mechanism of kernel attention with linear complexity is proposed to alleviate the large computational demand in attention. Based on kernel attention and channel attention, we integrate local feature maps extracted by ResNet-50 with their corresponding global dependencies and reweight interdependent channel maps adaptively. Numerical experiments on two large-scale fine resolution remote sensing datasets demonstrate the superior performance of the proposed MANet.

show abstract

Section: Quantitative Comparison Diverse Methodsmentioning

confidence: 99%

Section: ) Comparison With Simplified Dot-product Attentionmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Multiattention Network for Semantic Segmentation of Fine-Resolution Remote Sensing Images

Zheng

Zhang

et al. 2022

IEEE Trans. Geosci. Remote Sensing

Self Cite

126

View full text Add to dashboard Cite

show abstract

“…Recently, there has been an effort to reduce the memory footprint of the attention mechanism by introducing the concept of Linear attention [20]. This ideas was soon extended to 2D for computer vision problems [21]. In addition to these, the attention used in the recently introduced Visual Attention Transformer [22], which also helps in reducing the memory footprint for computer vision tasks.…”

Section: Related Work 1on Attentionmentioning

confidence: 99%

Looking for Change? Roll the Dice and Demand Attention

2021

View full text Add to dashboard Cite

Change detection, i.e., the identification per pixel of changes for some classes of interest from a set of bi-temporal co-registered images, is a fundamental task in the field of remote sensing. It remains challenging due to unrelated forms of change that appear at different times in input images. Here, we propose a deep learning framework for the task of semantic change detection in very high-resolution aerial images. Our framework consists of a new loss function, a new attention module, new feature extraction building blocks, and a new backbone architecture that is tailored for the task of semantic change detection. Specifically, we define a new form of set similarity that is based on an iterative evaluation of a variant of the Dice coefficient. We use this similarity metric to define a new loss function as well as a new, memory efficient, spatial and channel convolution Attention layer: the FracTAL. We introduce two new efficient self-contained feature extraction convolution units: the CEECNet and FracTALResNet units. Further, we propose a new encoder/decoder scheme, a network macro-topology, that is tailored for the task of change detection. The key insight in our approach is to facilitate the use of relative attention between two convolution layers in order to fuse them. We validate our approach by showing excellent performance and achieving state-of-the-art scores (F1 and Intersection over Union-hereafter IoU) on two building change detection datasets, namely, the LEVIRCD (F1: 0.918, IoU: 0.848) and the WHU (F1: 0.938, IoU: 0.882) datasets.

show abstract

“…This has led to the study of convolutional neural networks and the development of semantic segmentation for aerial imagery [17]. There are scientific publications comparing these methods with more recent semantic segmentation architectures in satellite imagery [5,18].…”

Section: Introductionmentioning

confidence: 99%

Evaluation of Semantic Segmentation Methods for Land Use with Spectral Imaging Using Sentinel-2 and PNOA Imagery

et al. 2021

View full text Add to dashboard Cite

Land use classification using aerial imagery can be complex. Characteristics such as ground sampling distance, resolution, number of bands and the information these bands convey are the keys to its accuracy. Random Forest is the most widely used approach but better and more modern alternatives do exist. In this paper, state-of-the-art methods are evaluated, consisting of semantic segmentation networks such as UNet and DeepLabV3+. In addition, two datasets based on aircraft and satellite imagery are generated as a new state of the art to test land use classification. These datasets, called UOPNOA and UOS2, are publicly available. In this work, the performance of these networks and the two datasets generated are evaluated. This paper demonstrates that ground sampling distance is the most important factor in obtaining good semantic segmentation results, but a suitable number of bands can be as important. This proves that both aircraft and satellite imagery can produce good results, although for different reasons. Finally, cost performance for an inference prototype is evaluated, comparing various Microsoft Azure architectures. The evaluation concludes that using a GPU is unnecessarily costly for deployment. A GPU need only be used for training.

show abstract

Multistage Attention ResU-Net for Semantic Segmentation of Fine-Resolution Remote Sensing Images

Cited by 91 publications

References 14 publications

Multiattention Network for Semantic Segmentation of Fine-Resolution Remote Sensing Images

Multiattention Network for Semantic Segmentation of Fine-Resolution Remote Sensing Images

Looking for Change? Roll the Dice and Demand Attention

Evaluation of Semantic Segmentation Methods for Land Use with Spectral Imaging Using Sentinel-2 and PNOA Imagery

Contact Info

Product

Resources

About