Sovereign 8s.is mirror of standardgalactic/podosome-segmentation
- Jupyter Notebook 98.7%
- Python 1.3%
| Filename | Latest commit message | Latest commit date |
|---|---|---|
|
|
||
| data | ||
| .gitignore | ||
| LICENSE | ||
| models.py | ||
| podosome-segmentation.ipynb | ||
| readme.txt | ||
| weight.pt | ||
Demo data and pretrained model weights for podosome image segmentation. -------------------------------------------------- 1. Contents -------------------------------------------------- - readme.txt This file. Instructions for setup and usage. - podosome-segmentation.ipynb Jupyter/Colab notebook for demonstration of segmentation - models.py Definition of segmentation model - weight.pt Trained weights of segmentation model - data/imgs.tif Example images for demonstration and testing purposes. -------------------------------------------------- 2. System Requirements -------------------------------------------------- Tested environment:: - Google Colab (Python 3.12.12) - PyTorch version 2.9.0 Required Python packages: - torch - matplotlib - numpy - tifffile Hardware: - GPU is recommended for faster execution, but CPU-only execution is also possible. -------------------------------------------------- 3. Installation -------------------------------------------------- To run the notebook in Google Colab: Upload the entire "podosome-segmentation" folder to your Google Drive under the "Colab Notebooks" directory. Installation time: < 10 seconds on Google Colab (required packages are automatically installed upon notebook execution) -------------------------------------------------- 4. Demo and Running the Code -------------------------------------------------- 1. Open the notebook file (podosome-segmentation.ipynb). 2. Run all cells in the notebook sequentially. The notebook produces: - Segmentation mask saved as a TIFF file in the "results" directory - Visualization of segmentation result -------------------------------------------------- 5. Expected Runtime -------------------------------------------------- - Approximately 10 seconds for the demo data on Google Colab with GPU -------------------------------------------------- 6. Using the Code with Your Own Data -------------------------------------------------- To apply the method to custom data: 1. Replace the example images in data with your own podosome images. 2. Update the input path in the notebook. 3. Run all cells. -------------------------------------------------- 7. License -------------------------------------------------- The code and trained weights are released under the GNU General Public License v3.0.