CellDetection 0.2.3

Cell Detection with PyTorch.

Cell Detection

⭐ Showcase

Nuclei of U2OS cells in a chemical screen

https://bbbc.broadinstitute.org/BBBC039 (CC0)

P. vivax (malaria) infected human blood

https://bbbc.broadinstitute.org/BBBC041 (CC BY-NC-SA 3.0)

🛠 Install

Make sure you have PyTorch installed.

PyPI

pip install -U celldetection

GitHub

pip install git+https://github.com/FZJ-INM1-BDA/celldetection.git

💡 How to train

Here you can see some examples of how to train a detection model. The examples already include toy data, so you can get started right away.

• Train your own model
• Train a model with multiple object classes

🔬 Models

from celldetection import models

Contour Proposal Networks:

• models.CpnU22
• models.CpnSlimU22
• models.CpnWideU22
• models.CpnResNet18FPN
• models.CpnResNet34FPN
• models.CpnResNet50FPN
• models.CpnResNet101FPN
• models.CpnResNet152FPN
• models.CpnResNeXt50FPN
• models.CpnResNeXt101FPN
• models.CpnResNeXt152FPN
• models.CpnWideResNet50FPN
• models.CpnWideResNet101FPN
• models.CpnMobileNetV3SmallFPN
• models.CpnMobileNetV3LargeFPN
• models.CPN

U-Nets:

• models.U22
• models.SlimU22
• models.WideU22
• models.U17
• models.U12
• models.UNetEncoder
• models.UNet

Feature Pyramid Networks:

• models.ResNet18FPN
• models.ResNet34FPN
• models.ResNet50FPN
• models.ResNet101FPN
• models.ResNet152FPN
• models.ResNeXt50FPN
• models.ResNeXt101FPN
• models.ResNeXt152FPN
• models.WideResNet50FPN
• models.WideResNet101FPN
• models.MobileNetV3SmallFPN
• models.MobileNetV3LargeFPN
• models.FPN

Residual Networks:

• models.ResNet18
• models.ResNet34
• models.ResNet50
• models.ResNet101
• models.ResNet152
• models.ResNeXt50_32x4d
• models.ResNeXt101_32x8d
• models.ResNeXt152_32x8d
• models.WideResNet50_2
• models.WideResNet101_2

Mobile Networks:

• models.MobileNetV3Small
• models.MobileNetV3Large

📝 Citing

@misc{upschulte2021contour,
      title={Contour Proposal Networks for Biomedical Instance Segmentation}, 
      author={Eric Upschulte and Stefan Harmeling and Katrin Amunts and Timo Dickscheid},
      year={2021},
      eprint={2104.03393},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

🔗 Links

• Abstract (arXiv)
• PDF (arXiv)
• Bibtex (arXiv)
• PyPI