GMDPUnet Model-Associated High-Resolution Optical Remote Sensing Semantic Segmentation Dataset

This dataset supports the research \GMDPUnet: A Global Modeling and Detail Perception Fusion Network for High-Resolution Remote Sensing Image Segmentation\, constructed by selecting data from the Vaihingen and Potsdam high-resolution optical remote sensing datasets.Along with the original pixel-level annotations for 6 land cover categories (e.g., buildings, vegetation, roads). The dataset can be directly used for training and validation of high-resolution remote sensing image semantic segmentation models (e.g., GMDPUnet), suitable for performance testing of global-local feature fusion networks, and provides standardized data support for comparative experiments of related segmentation algorithms.