ZephOD

Object/feature detection model selector algorithm. Current user version: v0.2.0.

Getting Started

1. Clone git repository:

git clone https://github.com/venkatachalamlab/zephod.git

2. Checkout the current version:

git checkout v0.2.0

Use the following command to see what's new in the most recent tagged version:

git show v0.2.0

3. Make sure that following Python libraries are installed (prefer conda over pip):

   • docopt
   • h5py
   • matplotlib
   • numpy
   • opencv
   • pandas
   • pathlib
   • pytorch
   • scipy
   • scikit-image
   • tqdm

4. Install (development mode):

(base) zephod> python setup.py develop

5. Train:

train_zephod --dataset=. --model=<yourmodelname> [options]

6. Run:

zephod --dataset=. --model=<yourmodelname> [options]

Current default channels

• Richardson-Lucy Deconvolution
• 0-255 Look-up Table
• Thresholding (low=50, high=200)
• Sharpening ([-3, 16, -3] kernel)

Currently available pretrained models

• celegans: pan-neuronal fluorescence image of freely-moving C. elegans worms