9 February 2026 in IEEE Transactions on Geoscience and Remote Sensing
Building change detection (BCD) using high-resolution satellite data is critical in urban surveillance, disaster management, and environmental assessments. However, issues such as resolution inconsistencies, spectrum fluctuation, and urban heterogeneity frequently restrict the reliability of current approaches. We proposed a novel Resolution-Standardized Change Detection Network (ReSCD-Net) to address these problems. The framework features an Adaptive Resolution Normalizer (ARN) for dynamic resolution standardization, a Resolution-Aware Segmentation Core (RASC) with a Hierarchical Spatial Extractor (HSE) for semantic representation at various scales, and a Contextual Refinement Decoder (CRD) for precise boundary reconstruction. Multi-scale characteristics are further enhanced by Atrous Spatial Pyramid Pooling (ASPP), allowing for robust contextual reasoning and detailed edge preservation. In addition, a Balanced Structural Loss (BSL) function is added to improve the trade-off between precision and recall while suppressing spurious changes caused by shadows, seasonal variations, and sensor noise. Three benchmark datasets covering various architectural and geographic circumstances (EGY-BCD, LEVIR-CD+, and DSIFN-CD) were used for the experiments. Achieving an F1-score of 91.09% and a mIoU of 91.12% on the EGY-BCD dataset, ReSCD-Net outperformed recent CNN- and Transformer-based baselines. With an F1-score of 93.06% and a mIoU of 90.83% on LEVIR-CD+, the model surpassed transformer-enhanced techniques while preserving a lower computing complexity. ReSCD-Net's balanced precision and recall scores on DSIFN-CD further validated its capacity for generalization in diverse urban and rural settings.
19 May 2026 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Zenith wet delay (ZWD) is one of the major sources of errors in the Global Navigation Satellite System (GNSS) positioning. Existing machine learning (ML)-based ZWD models have limitations in their generalization capabilities and in modeling the high spatiotemporal fluctuations in the atmospheric water vapor. To address these limitations, we introduce an advanced ensemble technique that integrates several base ML models through a Self-Supervised Uncertainty-Based Stacking Ensemble (SUSE). The proposed model enhances the accuracy of modeling and prediction of ZWD for China and its neighboring areas. The results demonstrate that our models, SUSE, outperform other ML, (DL), empirical (GPT3), and traditional models (Askne & Nordius (A&N) and Saastamoinen). In terms of RMSE, our blind model SUSE_A achieves internal and external improvements of 24.36% and 24.24% compared to GPT3. Additionally, SUSE_A outperforms conventional models that use meteorological data (A&N and Saastamoinen) and has comparable performance with other ML and DL models that use meteorological parameters. When surface meteorological data were available, the SUSE_B model reduced the RMSE by 34.48% and 47.89% with respect to A&N and Saastamoinen, respectively. Temporal and spatial analysis reveal the superior accuracy and stability of the SUSE model over other models examined in this research. These results indicate the promising prospects of the SUSE model for GNSS applications.