npsam 2.0.8

NP-SAM is an easy-to-use segmentation and analysis tool for nanoparticles and more.

NP-SAM

Introduction

In this project we propose an easily implementable workflow for a fast, accurate and seamless experience of segmentation of nanoparticles.

The project's experience can be significantly enhanced with the presence of a CUDA-compatible device; alternatively, Google Colab can be utilized if such a device is not accessible. For a quick access to the program and a CUDA-GPU try our Google Colab notebook.

Installation

Create a new conda environment called npsam and activate it.

conda create -n npsam python=3.10
conda activate npsam

Install PyTorch. NP-SAM has been tested with Pytorch 2.1.2 and CUDA 11.8.

Then install NP-SAM, and make a static link to the ipykernel

pip install npsam
python -m ipykernel install --user --name npsam --display-name npsam

Get started

In the npsam environment execute jupyter lab in the terminal. This will launch jupyterlab. Try out one of the example notebooks on our GitLab.

Citation

@article{NPSAM,
   author = {Larsen, Rasmus and Villadsen, Torben L. and Mathiesen, Jette K. and Jensen, Kirsten M. Ø and Bøjesen, Espen D.},
   title = {NP-SAM: Implementing the Segment Anything Model for Easy Nanoparticle Segmentation in Electron Microscopy Images},
   journal = {ChemRxiv},
   DOI = {10.26434/chemrxiv-2023-k73qz-v2},
   year = {2023},
   type = {Journal Article}
}

Acknowledgment

This repo benefits from Meta's Segment Anything and FastSAM. Thanks for their great work.