CM3Mamba: Preprocessed multisource remote sensing datasets and model code

This deposit contains preprocessed input datasets, experimental results, and supporting code for CM3Mamba-based multisource land cover classification. It accompanies the manuscript "CM3Mamba: Robust multisource land cover classification via three-slot state-space fusion and shared pure-read training" by Jingbo Zhou, Shuran Jia, Shaoqun Qi, Dousheng Zhang, and Zhenping Jiang. The prepared datasets include 11 × 11 training, validation, and test patches and labels for Houston 2013, Augsburg, and MUUFL Gulfport, together with fixed spatial split coordinates, sample ordering, evaluation subset definitions, and the normalization and PCA parameters used in preprocessing. HSI inputs contain 20 principal components. Auxiliary inputs comprise one LiDAR channel for Houston, four SAR channels and one DSM channel for Augsburg, and two LiDAR channels for MUUFL. Labels use zero-based class indices. The model code archive provides the proposed CM3Mamba model (configuration A6), its training entry point, environment requirements, and examples of fused and unimodal inference. The implementation is also available at https://github.com/nanhe-cmd/CM3Mamba. The accompanying analysis archive provides per-seed comparison and ablation metrics, aggregate kernel-control results, degradation and conditional-following results, inference-efficiency measurements, and observation statistics and spatial visualization inputs. It also includes experiment entry points and scripts for statistical analysis and numerical figure generation. A configuration mapping relates experiment script identifiers to those used in the manuscript. The README files and data dictionary describe file contents, metric definitions, statistical aggregation, software dependencies, and usage instructions; dataset_metadata.json provides machine-readable descriptions of the prepared datasets. Original data sources are the IEEE GRSS 2013 Data Fusion Contest (https://www.grss-ieee.org/community/technical-committees/2013-ieee-grss-data-fusion-contest/), the S2FL Augsburg benchmark (https://github.com/danfenghong/ISPRS_S2FL), and MUUFL Gulfport (https://github.com/GatorSense/MUUFLGulfport). Author-generated numerical results, documentation, and split descriptions are available under the Creative Commons Attribution 4.0 International license (CC BY 4.0), to the extent of the authors' rights. The CM3Mamba model code and the new analysis scripts are distributed under their included MIT licenses. Third-party observations, labels, software, and embedded source fragments retain their original providers' rights, license terms, and attribution requirements. Reuse and redistribution of source observations and derived patches follow the applicable provider terms.