This repository provides the code we have been using during our project development for Amharic language Word level text recognition using diffrent deep learning models. The code is provided as jupyter nootebook file and you can run it on google colab with GPU seeting on. The provided gode is very helpful for other HRTR and OCR researchers.
Please use our dataset provided on this link https://sites.google.com/view/hawdb-v1 and put the dataset on your google drive and mount it to your google colab if you are working on that. Or you can put it on your local drive if you run on your local GPU. Follow the details of the code provided on the jupyter notebook to train, validate and test the recognition models.
Do not forget to cite our paper: https://link.springer.com/article/10.1007/s42452-021-04742-x Contact US: [email protected]