Sanaj2060 / BTP2020

Tesseract and Transformer

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Manipuri Language - Tesseract OCR and NMT

First objective is to train a OCR model to recognise Manipuri words using Tesseract as backend engine. Second is to develop a transformer model for NMT which will be used in transliteration from Bengali script to Meitei Mayek. We also designed and develop the GUI feature for non-developers usage.

Steps to install

  1. Clone the repository locally.
  2. Install Tesseract-4.x.x version and its dependencies.
  3. Move the dataset/tesseract/mni.traineddata to tessdata folder.
  4. pip install -r requirements.txt
  5. ./manipuriOCR.sh
  6. Happy Hacking