This repository contains the code for the CS7NS1 Scalable Computing project.
The model and training are living in the crnn-ctc.ipynb
notebook.
To run the prediction on Raspberry Pi, use predict_lite.py
. Pre-trained model weights are available in the releases on GitHub.
Input images files should be placed in images
folder. Additionally, a images.csv
file should contain a list of input filenames.
The output will be written to output_lite.csv
.
TODO