CellSAM (ONNX)
This repository contains the ONNX export of the CellSAM model for cell segmentation.
⚠️ LICENSE WARNING: NON-COMMERCIAL / ACADEMIC USE ONLY ⚠️
This model is derived from the original CellSAM work and is subject to a Modified Apache License 2.0.
Grant of Copyright License: ... each Contributor hereby grants to You a non-commercial, academic perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable copyright license...
See the LICENSE file for full terms. Commercial use requires contacting the original authors (vanvalenlab@gmail.com).
Usage with cellsam-rs
This model is designed for high-performance inference using cellsam-rs.
# Install cellsam-rs (example)
cargo install cellsam-rs
# Run inference
cellsam-rs infer \
--input image.tif \
--model models/cellsam-onnx \
--output mask.tif
Model Details
- Original Repository: vanvalenlab/cellSAM
- Architecture: SAM-based (Segment Anything Model) specialized for cell segmentation.
- Input: Phase contrast + Fluorescence images (2-channel).
- Format: ONNX (Opset 17).
Citation
If you use this model, please cite the original CellSAM paper:
@article{Israel2024,
title={CellSAM: A Foundation Model for Cell Segmentation},
author={Israel, Uri and others},
journal={Nature Methods},
year={2024}
}