Haiyang Yu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Haiyang Yu has a total of 42 co-authors. Here are the top co-authors by number of shared publications:
- Pengao Li
Nanjing University of Information Science and Technology
5 shared publications - Ping Zhang
Henan Polytechnic University
4 shared publications - Bingqian Ji
School of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo 454000, China
2 shared publications - Fang Niu
School of Marxism Studies, Henan Polytechnic University, Jiaozuo 454003, China
2 shared publications Pengao Li
School of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo, China
2 shared publications- Qi Liu
School of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo 454000, China
2 shared publications
Latest Publications
- Open Access
- 12 September 2026
- Published by MDPI AG in Big Data and Cognitive Computing
- Vol. 10 (9) , 315
VFR-MISE: An Asymmetric Dual-Evidence Verification–Fallback Framework for Few-Shot Stressor Extraction from Social Media
Few-shot stressor extraction from temporally evolving social-media text remains challenging, primarily because of unstable span boundaries and the imperfect correspondence between structural confidence and task-semantic validity. We propose VFR-MISE, an asymmetric dual-evidence verification–fallback framework that extends the temporal few-shot knowledge-inheritance paradigm of MISE. The framework employs DeBERTa-v3-base for candidate generation, uses a guideline-conditioned verifier to assess candidate-level semantic validity, and applies BIOES emission evidence only to selectively recover high-confidence candidates rejected by the verifier. In a backbone comparison involving 20 random seeds, the DeBERTa-based candidate generator achieves a higher Exact-span F1 than its RoBERTa-based counterpart (0.7215 vs. 0.6942; +2.72 percentage points). The corresponding complete A3 systems achieve F1 scores of 0.7248 and 0.6977 (+2.70 percentage points; exact sign-flip test, p = 0.0000458). With the DeBERTa backbone fixed, the complete VFR-MISE framework improves the mean cross-shot Exact-span F1 by 0.33 percentage points over A0. Although this gain is modest, the overall results still provide directional support for the verification–fallback mechanism and indicate a relatively consistent refinement of the precision–recall trade-off. Computational cost analysis further shows that the full DeBERTa verifier introduces substantial additional computational overhead. In a matched benchmark, replacing the DeBERTa verifier with MiniLM reduces end-to-end inference time by 26.28% at a batch size of 16, suggesting that MiniLM is a promising candidate for lightweight verification. Overall, within the current experimental setting, VFR-MISE provides an auditable framework for stressor extraction that integrates semantic verification with constrained structural recovery.
- 12 September 2026
- Published by MDPI AG in Big Data and Cognitive Computing
- Vol. 10 (9) , 315
Open AccessVFR-MISE: An Asymmetric Dual-Evidence Verification–Fallback Framework for Few-Shot Stressor Extraction from Social Media
Few-shot stressor extraction from temporally evolving social-media text remains challenging, primarily because of unstable span boundaries and the imperfect correspondence between structural confidence and task-semantic validity. We propose VFR-MISE, an asymmetric dual-evidence verification–fallback framework that extends the temporal few-shot knowledge-inheritance paradigm of MISE. The framework employs DeBERTa-v3-base for candidate generation, uses a guideline-conditioned verifier to assess candidate-level semantic validity, and applies BIOES emission evidence only to selectively recover high-confidence candidates rejected by the verifier. In a backbone comparison involving 20 random seeds, the DeBERTa-based candidate generator achieves a higher Exact-span F1 than its RoBERTa-based counterpart (0.7215 vs. 0.6942; +2.72 percentage points). The corresponding complete A3 systems achieve F1 scores of 0.7248 and 0.6977 (+2.70 percentage points; exact sign-flip test, p = 0.0000458). With the DeBERTa backbone fixed, the complete VFR-MISE framework improves the mean cross-shot Exact-span F1 by 0.33 percentage points over A0. Although this gain is modest, the overall results still provide directional support for the verification–fallback mechanism and indicate a relatively consistent refinement of the precision–recall trade-off. Computational cost analysis further shows that the full DeBERTa verifier introduces substantial additional computational overhead. In a matched benchmark, replacing the DeBERTa verifier with MiniLM reduces end-to-end inference time by 26.28% at a batch size of 16, suggesting that MiniLM is a promising candidate for lightweight verification. Overall, within the current experimental setting, VFR-MISE provides an auditable framework for stressor extraction that integrates semantic verification with constrained structural recovery.
- Open Access
- 7 September 2026
- Published by MDPI AG in Preprints.org
VFR-MISE: An Asymmetric Dual-Evidence Verification–Fallback Framework for Few-Shot Stressor Extraction from Social Media
Few-shot stressor extraction from temporally shifting social-media text is challenged by unstable span boundaries and by a mismatch between structural confidence and task-semantic validity. We propose VFR-MISE, an asymmetric dual-evidence verification–fallback framework. VFR-MISE retains the temporal few-shot and knowledge-inheritance paradigm of MISE, strengthens candidate generation with DeBERTa-v3-base, and constructs two candidate-level signals: a guideline-conditioned verifier for semantic evidence and BIOES emission aggregation for structural evidence. Semantic evidence governs candidate admission; structural evidence is consulted only in the rejected region for constrained recovery, after which deterministic span merging produces the final output. In controlled reruns using identical data splits, training budgets, and Exact-span evaluation, replacing RoBERTa-base with DeBERTa-v3-base increases cross-shot F1 from 0.6841 to 0.7236 (+3.95 percentage points). With the DeBERTa-v3 backbone fixed, the complete VFR-MISE reaches 0.7248. These results suggest that stronger candidate representations and asymmetric dual-evidence decision refinement provide complementary benefits, improving the precision–recall trade-off while retaining an auditable decision process.
- 7 September 2026
- Published by MDPI AG in Preprints.org
Open AccessVFR-MISE: An Asymmetric Dual-Evidence Verification–Fallback Framework for Few-Shot Stressor Extraction from Social Media
Few-shot stressor extraction from temporally shifting social-media text is challenged by unstable span boundaries and by a mismatch between structural confidence and task-semantic validity. We propose VFR-MISE, an asymmetric dual-evidence verification–fallback framework. VFR-MISE retains the temporal few-shot and knowledge-inheritance paradigm of MISE, strengthens candidate generation with DeBERTa-v3-base, and constructs two candidate-level signals: a guideline-conditioned verifier for semantic evidence and BIOES emission aggregation for structural evidence. Semantic evidence governs candidate admission; structural evidence is consulted only in the rejected region for constrained recovery, after which deterministic span merging produces the final output. In controlled reruns using identical data splits, training budgets, and Exact-span evaluation, replacing RoBERTa-base with DeBERTa-v3-base increases cross-shot F1 from 0.6841 to 0.7236 (+3.95 percentage points). With the DeBERTa-v3 backbone fixed, the complete VFR-MISE reaches 0.7248. These results suggest that stronger candidate representations and asymmetric dual-evidence decision refinement provide complementary benefits, improving the precision–recall trade-off while retaining an auditable decision process.
- Open Access
- 13 February 2026
- Published by MDPI AG in Remote Sensing
- Vol. 18 (4) , 585
SOBLE-Top5: A Stacking Ensemble Learning-Based Seasonal Downscaling Inversion Framework for Surface Soil Moisture Using Multi-Source Data
Surface soil moisture (SSM) serves as a critical indicator for regional water cycles, agricultural management, and drought monitoring. However, existing the SMAP data suffers from limited spatial resolution, making it challenging to meet the demands of large-scale, high-resolution applications. Taking Henan Province, located in east-central China with a continental monsoon climate and marked seasonal variability, as the study area, this research integrates multi-source data to develop a seasonal modeling strategy. Based on stacking ensemble learning, the SSM downscaling inversion model (SOBLE-Top5) is constructed. SHAP value attribution analysis is employed to reveal the primary drivers of seasonal dynamics. The results indicate: (1) The SSM exhibits distinct seasonal characteristics. Compared to the all-season modeling, the RMSE and R2 metrics significantly improve during spring and summer. The winter ET and RF models show an approximately 9–14% higher R2 and a 47–50% lower RMSE. (2) The SOBLE-Top5 strategy achieved up to a 4.65% higher R2 and a 21.22% lower RMSE compared to the optimal single base model. (3) Spatial variations in the SSM characteristics reveal stable performance during the winter. The spring saw slight SSM declines in the northern regions due to rising temperatures. The study area reached its annual low (3/m3) in May–June. Driven by flood season precipitation, July–August witnessed local increases exceeding 52%. The autumn exhibited a stable-then-rising trend with pronounced north–south gradient characteristics. (4) The SHAP analysis indicates that the winter SSM is primarily controlled by bulk density and clay content. The spring SSM is most influenced by LST, followed by bulk density. The summer and the autumn SSM are synergistically driven by multiple factors including elevation, temperature, and precipitation, with the summer precipitation exerting the most significant impact on instantaneous SSM variations.
- 13 February 2026
- Published by MDPI AG in Remote Sensing
- Vol. 18 (4) , 585
Open AccessSOBLE-Top5: A Stacking Ensemble Learning-Based Seasonal Downscaling Inversion Framework for Surface Soil Moisture Using Multi-Source Data
Surface soil moisture (SSM) serves as a critical indicator for regional water cycles, agricultural management, and drought monitoring. However, existing the SMAP data suffers from limited spatial resolution, making it challenging to meet the demands of large-scale, high-resolution applications. Taking Henan Province, located in east-central China with a continental monsoon climate and marked seasonal variability, as the study area, this research integrates multi-source data to develop a seasonal modeling strategy. Based on stacking ensemble learning, the SSM downscaling inversion model (SOBLE-Top5) is constructed. SHAP value attribution analysis is employed to reveal the primary drivers of seasonal dynamics. The results indicate: (1) The SSM exhibits distinct seasonal characteristics. Compared to the all-season modeling, the RMSE and R2 metrics significantly improve during spring and summer. The winter ET and RF models show an approximately 9–14% higher R2 and a 47–50% lower RMSE. (2) The SOBLE-Top5 strategy achieved up to a 4.65% higher R2 and a 21.22% lower RMSE compared to the optimal single base model. (3) Spatial variations in the SSM characteristics reveal stable performance during the winter. The spring saw slight SSM declines in the northern regions due to rising temperatures. The study area reached its annual low (3/m3) in May–June. Driven by flood season precipitation, July–August witnessed local increases exceeding 52%. The autumn exhibited a stable-then-rising trend with pronounced north–south gradient characteristics. (4) The SHAP analysis indicates that the winter SSM is primarily controlled by bulk density and clay content. The spring SSM is most influenced by LST, followed by bulk density. The summer and the autumn SSM are synergistically driven by multiple factors including elevation, temperature, and precipitation, with the summer precipitation exerting the most significant impact on instantaneous SSM variations.
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19, 27647-27663
HRN: A Hybrid Recall Network for Crop Classification Using Multisource Time-Series Remote Sensing Data
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19, 27647-27663
Open AccessHRN: A Hybrid Recall Network for Crop Classification Using Multisource Time-Series Remote Sensing Data
- Open Access
- 11 December 2025
- Published by MDPI AG in Remote Sensing
- Vol. 17 (24) , 3996
MFEF-YOLO: A Multi-Scale Feature Extraction and Fusion Network for Small Object Detection in Aerial Imagery over Open Water
Current object detection using UAV platforms in open water faces challenges such as low detection accuracy, limited storage, and constrained computational capabilities. To address these issues, we propose MFEF-YOLO, a small object detection network based on multi-scale feature extraction and fusion. First, we introduce a Dual-Branch Spatial Pyramid Pooling Fast (DBSPPF) module in the backbone network to replace the original SPPF module, while integrating ODConv and C3k2 modules to collectively enhance feature extraction capabilities. Second, we improve small object detection by adding a P2 detection head and reduce model parameters by removing the P5 detection head. Finally, we design an Island-based Multi-scale Feature Fusion Network (IMFFNet) and employ a Coordinate-guided Multi-scale Feature Fusion Module (CMFFM) to strengthen contextual information and boost detection accuracy. We validate the effectiveness of MFEF-YOLO using the public dataset SeaDronesSee and our custom dataset TPDNV. Experimental results show that compared to the baseline model, mAP50 improves by 0.11 and 0.03 using the two datasets, respectively, while model parameters are reduced by 11.54%. Furthermore, DBSPPF and IMFFNet demonstrate superior performance in comparative studies with other methods, confirming their effectiveness. These improvements and outstanding performance make MFEF-YOLO particularly suitable for UAV-based object detection in open waters.
- 11 December 2025
- Published by MDPI AG in Remote Sensing
- Vol. 17 (24) , 3996
Open AccessMFEF-YOLO: A Multi-Scale Feature Extraction and Fusion Network for Small Object Detection in Aerial Imagery over Open Water
Current object detection using UAV platforms in open water faces challenges such as low detection accuracy, limited storage, and constrained computational capabilities. To address these issues, we propose MFEF-YOLO, a small object detection network based on multi-scale feature extraction and fusion. First, we introduce a Dual-Branch Spatial Pyramid Pooling Fast (DBSPPF) module in the backbone network to replace the original SPPF module, while integrating ODConv and C3k2 modules to collectively enhance feature extraction capabilities. Second, we improve small object detection by adding a P2 detection head and reduce model parameters by removing the P5 detection head. Finally, we design an Island-based Multi-scale Feature Fusion Network (IMFFNet) and employ a Coordinate-guided Multi-scale Feature Fusion Module (CMFFM) to strengthen contextual information and boost detection accuracy. We validate the effectiveness of MFEF-YOLO using the public dataset SeaDronesSee and our custom dataset TPDNV. Experimental results show that compared to the baseline model, mAP50 improves by 0.11 and 0.03 using the two datasets, respectively, while model parameters are reduced by 11.54%. Furthermore, DBSPPF and IMFFNet demonstrate superior performance in comparative studies with other methods, confirming their effectiveness. These improvements and outstanding performance make MFEF-YOLO particularly suitable for UAV-based object detection in open waters.
- Open Access
- 10 January 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-22
InSAR-YOLOv8 for wide-area landslide detection in InSAR measurements
InSAR monitoring technology is widely used in investigating landslide hazards. Leveraging object detection algorithms to quickly extract landslide information from Wide-Area InSAR measurements is of great significance. Our InSAR-YOLOv8, an algorithm that automatically detects landslides from InSAR measurements, addresses the low accuracy and suboptimal detection performance of existing network models. In this method, we first design and add a detection head specifically targeting small-scale objects. This improvement enhances the model’s ability to extract features across different scales and strengthens its capability to detect landslides of varying sizes. We also replace the original C2f module with the lighter C2f_Faster module to process information more efficiently, making the model lighter and more efficient. Finally, the SIoU loss function replaces the CIoU loss function to improve the bounding box regression and enhance detection accuracy. Our results show that the proposed algorithm achieves a 97.41% mAP50, a 66.47% mAP50:95, and a 92.06% F1 score on the InSAR landslide dataset, while reducing the number of parameters by 25%. Compared with YOLOv8 and other advanced models (YOLOvX, Faster R-CNN, etc.), our model exhibits distinct advantages and possesses a wider range of potential applications in InSAR measurement for landslide detection.
- 10 January 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-22
Open AccessInSAR-YOLOv8 for wide-area landslide detection in InSAR measurements
InSAR monitoring technology is widely used in investigating landslide hazards. Leveraging object detection algorithms to quickly extract landslide information from Wide-Area InSAR measurements is of great significance. Our InSAR-YOLOv8, an algorithm that automatically detects landslides from InSAR measurements, addresses the low accuracy and suboptimal detection performance of existing network models. In this method, we first design and add a detection head specifically targeting small-scale objects. This improvement enhances the model’s ability to extract features across different scales and strengthens its capability to detect landslides of varying sizes. We also replace the original C2f module with the lighter C2f_Faster module to process information more efficiently, making the model lighter and more efficient. Finally, the SIoU loss function replaces the CIoU loss function to improve the bounding box regression and enhance detection accuracy. Our results show that the proposed algorithm achieves a 97.41% mAP50, a 66.47% mAP50:95, and a 92.06% F1 score on the InSAR landslide dataset, while reducing the number of parameters by 25%. Compared with YOLOv8 and other advanced models (YOLOvX, Faster R-CNN, etc.), our model exhibits distinct advantages and possesses a wider range of potential applications in InSAR measurement for landslide detection.
- Open Access
- 13 November 2024
- Published by Frontiers Media SA in Frontiers in Sustainable Food Systems
- Vol. 8, 1433701
Enhancing sustainable Chinese cabbage production: a comparative analysis of multispectral image instance segmentation techniques
Accurate instance segmentation of individual crops is crucial for field management and crop monitoring in smart agriculture. To address the limitations of traditional remote sensing methods in individual crop analysis, this study proposes a novel instance segmentation approach combining UAVs with the YOLOv8-Seg model. The YOLOv8-Seg model supports independent segmentation masks and detection at different scales, utilizing Path Aggregation Feature Pyramid Networks (PAFPN) for multi-scale feature integration and optimizing sample matching through the Task-Aligned Assigner. We collected multispectral data of Chinese cabbage using UAVs and constructed a high-quality dataset via semi-automatic annotation with the Segment Anything Model (SAM). Using mAP as the evaluation metric, we compared YOLO series algorithms with other mainstream instance segmentation methods and analyzed model performance under different spectral band combinations and spatial resolutions. The results show that YOLOv8-Seg achieved 86.3% mAP under the RGB band and maintained high segmentation accuracy at lower spatial resolutions (1.33 ~ 1.14 cm/pixel), successfully extracting key metrics such as cabbage count and average leaf area. These findings highlight the potential of integrating UAV technology with advanced segmentation models for individual crop monitoring, supporting precision agriculture applications.
- 13 November 2024
- Published by Frontiers Media SA in Frontiers in Sustainable Food Systems
- Vol. 8, 1433701
Open AccessEnhancing sustainable Chinese cabbage production: a comparative analysis of multispectral image instance segmentation techniques
Accurate instance segmentation of individual crops is crucial for field management and crop monitoring in smart agriculture. To address the limitations of traditional remote sensing methods in individual crop analysis, this study proposes a novel instance segmentation approach combining UAVs with the YOLOv8-Seg model. The YOLOv8-Seg model supports independent segmentation masks and detection at different scales, utilizing Path Aggregation Feature Pyramid Networks (PAFPN) for multi-scale feature integration and optimizing sample matching through the Task-Aligned Assigner. We collected multispectral data of Chinese cabbage using UAVs and constructed a high-quality dataset via semi-automatic annotation with the Segment Anything Model (SAM). Using mAP as the evaluation metric, we compared YOLO series algorithms with other mainstream instance segmentation methods and analyzed model performance under different spectral band combinations and spatial resolutions. The results show that YOLOv8-Seg achieved 86.3% mAP under the RGB band and maintained high segmentation accuracy at lower spatial resolutions (1.33 ~ 1.14 cm/pixel), successfully extracting key metrics such as cabbage count and average leaf area. These findings highlight the potential of integrating UAV technology with advanced segmentation models for individual crop monitoring, supporting precision agriculture applications.
- Open Access
Research on Segmentation Method of Maize Seedling Plant Instances Based on UAV Multispectral Remote Sensing Images
The accurate instance segmentation of individual crop plants is crucial for achieving a high-throughput phenotypic analysis of seedlings and smart field management in agriculture. Current crop monitoring techniques employing remote sensing predominantly focus on population analysis, thereby lacking precise estimations for individual plants. This study concentrates on maize, a critical staple crop, and leverages multispectral remote sensing data sourced from unmanned aerial vehicles (UAVs). A large-scale SAM image segmentation model is employed to efficiently annotate maize plant instances, thereby constructing a dataset for maize seedling instance segmentation. The study evaluates the experimental accuracy of six instance segmentation algorithms: Mask R-CNN, Cascade Mask R-CNN, PointRend, YOLOv5, Mask Scoring R-CNN, and YOLOv8, employing various combinations of multispectral bands for a comparative analysis. The experimental findings indicate that the YOLOv8 model exhibits exceptional segmentation accuracy, notably in the NRG band, with bbox_mAP50 and segm_mAP50 accuracies reaching 95.2% and 94%, respectively, surpassing other models. Furthermore, YOLOv8 demonstrates robust performance in generalization experiments, indicating its adaptability across diverse environments and conditions. Additionally, this study simulates and analyzes the impact of different resolutions on the model’s segmentation accuracy. The findings reveal that the YOLOv8 model sustains high segmentation accuracy even at reduced resolutions (1.333 cm/px), meeting the phenotypic analysis and field management criteria.
Open AccessResearch on Segmentation Method of Maize Seedling Plant Instances Based on UAV Multispectral Remote Sensing Images
The accurate instance segmentation of individual crop plants is crucial for achieving a high-throughput phenotypic analysis of seedlings and smart field management in agriculture. Current crop monitoring techniques employing remote sensing predominantly focus on population analysis, thereby lacking precise estimations for individual plants. This study concentrates on maize, a critical staple crop, and leverages multispectral remote sensing data sourced from unmanned aerial vehicles (UAVs). A large-scale SAM image segmentation model is employed to efficiently annotate maize plant instances, thereby constructing a dataset for maize seedling instance segmentation. The study evaluates the experimental accuracy of six instance segmentation algorithms: Mask R-CNN, Cascade Mask R-CNN, PointRend, YOLOv5, Mask Scoring R-CNN, and YOLOv8, employing various combinations of multispectral bands for a comparative analysis. The experimental findings indicate that the YOLOv8 model exhibits exceptional segmentation accuracy, notably in the NRG band, with bbox_mAP50 and segm_mAP50 accuracies reaching 95.2% and 94%, respectively, surpassing other models. Furthermore, YOLOv8 demonstrates robust performance in generalization experiments, indicating its adaptability across diverse environments and conditions. Additionally, this study simulates and analyzes the impact of different resolutions on the model’s segmentation accuracy. The findings reveal that the YOLOv8 model sustains high segmentation accuracy even at reduced resolutions (1.333 cm/px), meeting the phenotypic analysis and field management criteria.
- Open Access
Flood Detection in Polarimetric SAR Data Using Deformable Convolutional Vision Model
Floods represent a significant natural hazard with the potential to inflict substantial damage on human society. The swift and precise delineation of flood extents is of paramount importance for effectively supporting flood response and disaster relief efforts. In comparison to optical sensors, Synthetic Aperture Radar (SAR) sensor data acquisition exhibits superior capabilities, finding extensive application in flood detection research. Nonetheless, current methodologies exhibit limited accuracy in flood boundary detection, leading to elevated instances of both false positives and false negatives, particularly in the detection of smaller-scale features. In this study, we proposed an advanced flood detection method called FWSARNet, which leveraged a deformable convolutional visual model with Sentinel-1 SAR images as its primary data source. This model centered around deformable convolutions as its fundamental operation and took inspiration from the structural merits of the Vision Transformer. Through the introduction of a modest number of supplementary parameters, it significantly extended the effective receptive field, enabling the comprehensive capture of intricate local details and spatial fluctuations within flood boundaries. Moreover, our model employed a multi-level feature map fusion strategy that amalgamated feature information from diverse hierarchical levels. This enhancement substantially augmented the model’s capability to encompass various scales and boost its discriminative power. To validate the effectiveness of the proposed model, experiments were conducted using the ETCI2021 dataset. The results demonstrated that the Intersection over Union (IoU) and mean Intersection over Union (mIoU) metrics for flood detection achieved impressive values of 80.10% and 88.47%, respectively. These results surpassed the performance of state-of-the-art (SOTA) models. Notably, in comparison to the best results documented on the official ETCI2021 dataset competition website, our proposed model in this paper exhibited a remarkable 3.29% improvement in flood prediction IoU. The experimental outcomes underscore the capability of the FWSARNet method outlined in this paper for flood detection using Synthetic Aperture Radar (SAR) data. This method notably enhances the accuracy of flood detection, providing essential technical and data support for real-world flood monitoring, prevention, and response efforts.
Open AccessFlood Detection in Polarimetric SAR Data Using Deformable Convolutional Vision Model
Floods represent a significant natural hazard with the potential to inflict substantial damage on human society. The swift and precise delineation of flood extents is of paramount importance for effectively supporting flood response and disaster relief efforts. In comparison to optical sensors, Synthetic Aperture Radar (SAR) sensor data acquisition exhibits superior capabilities, finding extensive application in flood detection research. Nonetheless, current methodologies exhibit limited accuracy in flood boundary detection, leading to elevated instances of both false positives and false negatives, particularly in the detection of smaller-scale features. In this study, we proposed an advanced flood detection method called FWSARNet, which leveraged a deformable convolutional visual model with Sentinel-1 SAR images as its primary data source. This model centered around deformable convolutions as its fundamental operation and took inspiration from the structural merits of the Vision Transformer. Through the introduction of a modest number of supplementary parameters, it significantly extended the effective receptive field, enabling the comprehensive capture of intricate local details and spatial fluctuations within flood boundaries. Moreover, our model employed a multi-level feature map fusion strategy that amalgamated feature information from diverse hierarchical levels. This enhancement substantially augmented the model’s capability to encompass various scales and boost its discriminative power. To validate the effectiveness of the proposed model, experiments were conducted using the ETCI2021 dataset. The results demonstrated that the Intersection over Union (IoU) and mean Intersection over Union (mIoU) metrics for flood detection achieved impressive values of 80.10% and 88.47%, respectively. These results surpassed the performance of state-of-the-art (SOTA) models. Notably, in comparison to the best results documented on the official ETCI2021 dataset competition website, our proposed model in this paper exhibited a remarkable 3.29% improvement in flood prediction IoU. The experimental outcomes underscore the capability of the FWSARNet method outlined in this paper for flood detection using Synthetic Aperture Radar (SAR) data. This method notably enhances the accuracy of flood detection, providing essential technical and data support for real-world flood monitoring, prevention, and response efforts.
- Open Access
- 12 November 2023
- Published by MDPI AG in Remote Sensing
- Vol. 15 (22) , 5331
TransHSI: A Hybrid CNN-Transformer Method for Disjoint Sample-Based Hyperspectral Image Classification
Hyperspectral images’ (HSIs) classification research has seen significant progress with the use of convolutional neural networks (CNNs) and Transformer blocks. However, these studies primarily incorporated Transformer blocks at the end of their network architectures. Due to significant differences between the spectral and spatial features in HSIs, the extraction of both global and local spectral–spatial features remains incomplete. To address this challenge, this paper introduces a novel method called TransHSI. This method incorporates a new spectral–spatial feature extraction module that leverages 3D CNNs to fuse Transformer to extract the local and global spectral features of HSIs, then combining 2D CNNs and Transformer to capture the local and global spatial features of HSIs comprehensively. Furthermore, a fusion module is proposed, which not only integrates the learned shallow and deep features of HSIs but also applies a semantic tokenizer to transform the fused features, enhancing the discriminative power of the features. This paper conducts experiments on three public datasets: Indian Pines, Pavia University, and Data Fusion Contest 2018. The training and test sets are selected based on a disjoint sampling strategy. We perform a comparative analysis with 11 traditional and advanced HSI classification algorithms. The experimental results demonstrate that the proposed method, TransHSI algorithm, achieves the highest overall accuracies and kappa coefficients, indicating a competitive performance.
- 12 November 2023
- Published by MDPI AG in Remote Sensing
- Vol. 15 (22) , 5331
Open AccessTransHSI: A Hybrid CNN-Transformer Method for Disjoint Sample-Based Hyperspectral Image Classification
Hyperspectral images’ (HSIs) classification research has seen significant progress with the use of convolutional neural networks (CNNs) and Transformer blocks. However, these studies primarily incorporated Transformer blocks at the end of their network architectures. Due to significant differences between the spectral and spatial features in HSIs, the extraction of both global and local spectral–spatial features remains incomplete. To address this challenge, this paper introduces a novel method called TransHSI. This method incorporates a new spectral–spatial feature extraction module that leverages 3D CNNs to fuse Transformer to extract the local and global spectral features of HSIs, then combining 2D CNNs and Transformer to capture the local and global spatial features of HSIs comprehensively. Furthermore, a fusion module is proposed, which not only integrates the learned shallow and deep features of HSIs but also applies a semantic tokenizer to transform the fused features, enhancing the discriminative power of the features. This paper conducts experiments on three public datasets: Indian Pines, Pavia University, and Data Fusion Contest 2018. The training and test sets are selected based on a disjoint sampling strategy. We perform a comparative analysis with 11 traditional and advanced HSI classification algorithms. The experimental results demonstrate that the proposed method, TransHSI algorithm, achieves the highest overall accuracies and kappa coefficients, indicating a competitive performance.