2026-10-14T14:42:01Z https://www.scidb.cn/oai

10.57760/sciencedb.00zkt 2026-08-31T10:16:38Z

2026-08-31 Experimental Code and Data for "Forest Cover Mapping from Landsat Imagery via Swin Transformer with Global‑Local Synergy" doi:10.57760/sciencedb.00zkt en This dataset provides the Python scripts and model files for implementing the proposed global‑local collaborative Swin Transformer framework, including the dual gate‑controlled local context module (DGLCM), deformable convolution, stripe pooling, and a multi‑level gated residual fusion mechanism. In addition, the experimental result data—such as accuracy metrics, confusion matrices, and the numerical values used to generate all figures—are also included. forest cover; remote sensing mapping; Swin Transformer; deformable convolution; strip pooling Jinyuan Zheng PUBLIC https://creativecommons.org/licenses/by/4.0/ dataset Science Data Bank