Collected molecules will appear here. Add from search or explore.
Provides a specialized deep learning approach for segmenting buildings in satellite imagery captured at high off-nadir angles (oblique views), addressing geometric distortion and resolution loss.
citations
0
co_authors
7
This is an academic reference implementation for a 2021 paper. While it addresses a specific pain point in remote sensing (off-nadir distortion), the project has zero stars and no community momentum. Since its publication, geospatial foundation models (like IBM/NASA Prithvi or Segment Anything variants) have largely superseded specialized single-task models for this type of imagery analysis.
TECH STACK
INTEGRATION
reference_implementation
READINESS