Mask2Former with Improved Query for Semantic Segmentation in Remote-Sensing Images

Guo, Shichen; Yang, Qi; Xiang, Shiming; Wang, Shuwen; Wang, Xuezhi

doi:10.3390/math12050765

Cited by 3 publications

(1 citation statement)

References 62 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In PSPNet, the pyramid pooling module is designed to aggregate contextual information of different scales [50]. DeepLabV3+ employs dilated convolution to enlarge the receptive field of filters and introduces feature pyramid network to combine features at different spatial resolutions [51]. All these methods concentrate on extracting multi-scale information for better semantic segmentation performance.…”

Section: Comparison With Other Modelsmentioning

confidence: 99%

Generating 10-Meter Resolution Land Use and Land Cover Products Using Historical Landsat Archive Based on Super Resolution Guided Semantic Segmentation Network

Wen,

Zhu,

Tian

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

Generating high-resolution land cover maps using relatively lower-resolution remote sensing images is of great importance for subtle analysis. However, the domain gap between real lower-resolution and synthetic images has not been permanently resolved. Furthermore, super-resolution information is not fully exploited in semantic segmentation models. By solving the aforementioned issues, a deeply fused super resolution guided semantic segmentation network using 30 m Landsat images is proposed. A large-scale dataset comprising 10 m Sentinel-2, 30 m Landsat-8 images, and 10 m European Space Agency (ESA) Land Cover Product is introduced, facilitating model training and evaluation across diverse real-world scenarios. The proposed Deeply Fused Super Resolution Guided Semantic Segmentation Network (DFSRSSN) combines a Super Resolution Module (SRResNet) and a Semantic Segmentation Module (CRFFNet). SRResNet enhances spatial resolution, while CRFFNet leverages super-resolution information for finer-grained land cover classification. Experimental results demonstrate the superior performance of the proposed method in five different testing datasets, achieving 68.17–83.29% and 39.55–75.92% for overall accuracy and kappa, respectively. When compared to ResUnet with up-sampling block, increases of 2.16–34.27% and 8.32–43.97% were observed for overall accuracy and kappa, respectively. Moreover, we proposed a relative drop rate of accuracy metrics to evaluate the transferability. The model exhibits improved spatial transferability, demonstrating its effectiveness in generating accurate land cover maps for different cities. Multi-temporal analysis reveals the potential of the proposed method for studying land cover and land use changes over time. In addition, a comparison of the state-of-the-art full semantic segmentation models indicates that spatial details are fully exploited and presented in semantic segmentation results by the proposed method.

show abstract

Section: Comparison With Other Modelsmentioning

confidence: 99%

Generating 10-Meter Resolution Land Use and Land Cover Products Using Historical Landsat Archive Based on Super Resolution Guided Semantic Segmentation Network

Wen,

Zhu,

Tian

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

show abstract

Quality Evaluation of Public Spaces in Traditional Villages: A Study Using Deep Learning and Panoramic Images

Meng,

Liu,

Zeng

et al. 2024

Land

View full text Add to dashboard Cite

In the context of rapid urbanization, public spaces in traditional villages face challenges such as material ageing, loss of characteristics, and functional decline. The scientific and objective assessment of the quality of these public spaces is crucial for the sustainable development of traditional villages. Panoramic images, as an important source of spatial data, combined with deep learning technology, can objectively quantify the characteristics of public spaces in traditional villages. However, existing research has paid insufficient attention to the evaluation of the quality of public spaces in traditional villages at the micro-scale, often relying on questionnaires and interviews, which makes it difficult to meet the needs of planning and construction. This study constructs an evaluation system for the quality of public spaces in traditional villages, taking national-level traditional villages in the Fangshan District of Beijing as an example, based on traditional field research, using deep learning and panoramic images to automatically extract the features of public spaces in traditional villages, using a combination of the Analytic Hierarchy Process (AHP) and Criteria Importance Through Intercriteria Correlation (CRITIC) methods to determine the weights of the indicators and applying the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) method to evaluate the quality of public spaces in traditional villages. The study found that the quality of public spaces in Nanjiao Village is Grade I; Shuiyu Village and Liulinshui Village, Grade III; and Heilongguan Village, Grade IV and that there is still much room for improvement in general. The evaluation results match well with the public’s subjective perceptions, with an R2 value of 0.832, proving that the constructed evaluation system has a high degree of accuracy. This study provides a scientific basis and an effective tool for the planning, design, and management of public spaces in traditional villages, which helps decision-makers better protect and utilize them.

show abstract

Improving Safety in High-Altitude Work: Semantic Segmentation of Safety Harnesses with CEMFormer

Zhou,

Liu

2024

Symmetry

View full text Add to dashboard Cite

The symmetry between production efficiency and safety is a crucial aspect of industrial operations. To enhance the identification of proper safety harness use by workers at height, this study introduces a machine vision approach as a substitute for manual supervision. By focusing on the safety rope that connects the worker to an anchor point, we propose a semantic segmentation mask annotation principle to evaluate proper harness use. We introduce CEMFormer, a novel semantic segmentation model utilizing ConvNeXt as the backbone, which surpasses the traditional ResNet in accuracy. Efficient Multi-Scale Attention (EMA) is incorporated to optimize channel weights and integrate spatial information. Mask2Former serves as the segmentation head, enhanced by Poly Loss for classification and Log-Cosh Dice Loss for mask loss, thereby improving training efficiency. Experimental results indicate that CEMFormer achieves a mean accuracy of 92.31%, surpassing the baseline and five state-of-the-art models. Ablation studies underscore the contribution of each component to the model’s accuracy, demonstrating the effectiveness of the proposed approach in ensuring worker safety.

show abstract

Mask2Former with Improved Query for Semantic Segmentation in Remote-Sensing Images

Cited by 3 publications

References 62 publications

Generating 10-Meter Resolution Land Use and Land Cover Products Using Historical Landsat Archive Based on Super Resolution Guided Semantic Segmentation Network

Generating 10-Meter Resolution Land Use and Land Cover Products Using Historical Landsat Archive Based on Super Resolution Guided Semantic Segmentation Network

Quality Evaluation of Public Spaces in Traditional Villages: A Study Using Deep Learning and Panoramic Images

Improving Safety in High-Altitude Work: Semantic Segmentation of Safety Harnesses with CEMFormer

Contact Info

Product

Resources

About