---
title: SinkSAM Net
emoji: 🌍
colorFrom: gray
colorTo: yellow
sdk: streamlit
sdk_version: 1.44.1
app_file: app.py
pinned: false
license: mit
short_description: Knowledge-Driven Self-Supervised Sinkhole Segmentation
tags:
  - object-detection
  - segmentation
  - remote-sensing
  - geoscience
---

This is a demo is a simplified version of the approach described in the paper, ["SinkSAM: A Monocular Depth-Guided SAM Framework for Automatic Sinkhole Segmentation
"](https://arxiv.org/abs/2410.01473)

```
@misc{rafaeli2024sinksammonoculardepthguidedsam,
      title={SinkSAM: A Monocular Depth-Guided SAM Framework for Automatic Sinkhole Segmentation}, 
      author={Osher Rafaeli and Tal Svoray and Ariel Nahlieli},
      year={2024},
      eprint={2410.01473},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2410.01473}, 
}
```