MAPUNet: Architecture Table for Mining Area Deformation Monitoring Model

A table detailing the architecture of the MAPUNet model, a multi-scale phase unwrapping method for satellite interferometry. The model was developed to monitor large gradient deformations in mining areas, with validation on simulated interferograms and a real case in Shanxi, China. The dataset was uploaded by Baojing Zhang on figshare in May 2026.

Use Cases

• Implementing the MAPUNet architecture based on the detailed structural parameters.
• Benchmarking phase unwrapping models against the reported SSIM performance metrics.
• Applying the model to monitor large-scale subsidence in mining areas based on the described methodology.
• Studying the integration of atrous spatial pyramid pooling (ASPP) and dual attention networks (DA-Net) for feature extraction.

Strengths

• The model demonstrated improved accuracy, with SSIM increasing by 0.298 compared to the original PUNet model under high-noise conditions.
• The dataset is openly licensed under CC-BY-4.0, allowing for reuse and modification.
• The method was validated on real data, achieving deformation inversion for a mining area in Shanxi with a measured maximum settlement of -9.3 cm within one month.

Limitations

• The dataset is very small (5.5 KB), suggesting it contains only a structural table, not the underlying training or validation data.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for certain analyses.

Provenance

Source

figshare

Collection Method

Likely generated as part of the research paper describing the MAPUNet model.

Freshness

Last updated 2026-05-26 17:39:12; freshness should be verified.

Geography

Method validated on a mining area in Shanxi, China.