Skip to content

Team & credits

Department of Artificial Intelligence & Data Science, Jawaharlal Nehru Engineering College, Chhatrapati Sambhaji Nagar.

  • Abhay Kale
  • Akash Chaudhari
  • Somesh Padsalge

With thanks to Dr. V. B. Malode for guidance and review throughout this work.

  • GAMUS: geometry-aware multi-modal semantic segmentation benchmark (nDSM + classes).
  • SynRS3D: synthetic remote-sensing 3D dataset.
  • IEEE GRSS Data Fusion Contest 2023 (Track 2) and 2019 (US3D).
  • IARPA Multi-View Stereo 3D Mapping benchmark: imagery courtesy of DigitalGlobe, approved for public release.
  • NEON AOP: National Ecological Observatory Network airborne LiDAR (canopy height model).
  • Copernicus DEM: produced using Copernicus WorldDEM-30 © DLR e.V. 2010–2014 and © Airbus Defence and Space GmbH 2014–2018, provided under COPERNICUS by the European Union and ESA; all rights reserved.
  • SRTM / NASADEM (NASA), Terrarium tiles (AWS Open Data), OpenStreetMap contributors (ODbL).
  • Cartosat-2S / 2E sample products: NRSC Bhoonidhi.
  • Screenshots marked © Google Earth are used for illustration only.

Packed training stores and the trained model are listed under Download the data.

PyTorch, Hugging Face Transformers (DINOv3), ONNX Runtime, FastAPI, Gradio, rasterio / GDAL / PROJ, React, three.js, React Three Fiber, Mantine, ECharts, and Astro Starlight for these docs.

MIT License. See LICENSE in the repository root.