Cheetah is a simple to use Python library to support the development of analysis workflows and real-time cybergenetic control platforms that combine microscopy imaging and environmental control. The toolkit relies on a U-Net convolutional neural network to perform cell segmentation and uses this with additional functions that can label, track and generate control signals based on this information. If you make use of Cheetah in your work we ask that you cite the following paper:
If you would like to use Cheetah for your own project, we recommend using Anaconda Python and creating a separate environment where all the necessary dependancies can be installed. This can be done after install Anaconda by running the following commands:
conda create -n py36 python=3.6
conda activate py36
conda install keras==2.0.8 matplotlib numpy scipy scikit-image scikit-learn
To test if this has worked successfully, it should now be possible to run the training step for our bacteria example in the examples/01_bacteria by running the command:
python 02_training.py
This should start up the training cycle using the example bacteria training and validation images provided as part of the package. We also recommend taking a look at some of the other examples provided and reading the associated publication to better understand how all aspects of the functionality can be used.
keras, tensorflow, scikit-learn, scikit-image, scipy, glob, numpy, matplotlib, json