Landslide deformation monitoring via InSAR is crucial for assessing the risk of hazards. Quick and accurate detection of active deformation zones is crucial for early warning and mitigation planning. While the application of deep learning has substantially improved the detection efficiency, several challenges still persist, such as poor multi-scale perception, blurred boundaries, and limited model generalization. This study proposes Hybrid-SegUFormer to address these limitations. The model integrates the SegFormer encoder’s efficient feature extraction with the U-Net decoder’s superior boundary restoration. It introduces a multi-scale fusion decoding mechanism to enhance context perception structurally and incorporates a self-distillation strategy to significantly improve generalization capability. Hybrid-SegUFormer achieves detection performance (98.79% accuracy, 80.05% F1-score) while demonstrating superior multi-scale adaptability (IoU degradation of only 6.99–8.83%) and strong cross-regional generalization capability. The synergistic integration of its core modules enables an optimal balance between precision and recall, making it particularly effective for complex landslide detection tasks. This study provides a new approach for intelligent interpretation of InSAR deformation in complex mountainous areas.

With advances in seafloor sensor data acquisition and deep-learning–based image interpretation, modern deep-sea resource exploration has progressed substantially; however, traditional semantic segmentation models remain challenged on seafloor optical mineral imagery due to pronounced multi-scale feature heterogeneity, object coalescence, and frequent missed detections of small-scale mineral targets. To this end, this study presents AttMSD-Uformer, a deep-learning semantic segmentation architecture that couples a dual-branch encoder with an attention-guided multi-scale decoder, and further introduces a semi-automated annotation pipeline that combines algorithmic recognition with expert-guided refinement to curate high-quality training datasets efficiently. Experimental results show that the proposed model delivers consistently strong segmentation performance across multiple seafloor mineral datasets and demonstrates robust cross-dataset generalization. Furthermore, this study systematically compared the predictive performance of four representative regression models using segmentation-derived image-level nodule coverage as the explanatory variable. The results indicate that the random-forest regressor is particularly well suited to the heterogeneous characteristics of seafloor polymetallic nodule imagery, excelling at modeling nonlinear and piecewise relationships. This methodology provides an efficient and reliable reference framework for intelligent quantitative assessment of seafloor resources. The code is available at https://github.com/jiahaoZhang-cugb/AttMSD-Uformer.