[[Arxiv-PDF-Link

In this paper, we proposed a convolutional architecture for the task of handwritten text recognition that utilizes only 1D convolutions, dense residual connections and a SE module. We also proposed a simple and effective data augmentation technique-’Tiling and Corruption (TACo)’ useful for OCR/HTR tasks. We have presented experimental study on components of Easter2.0 architecture including dense residual connections, normalization choices, SE module, TACo variations and few-shot training. Our work achieves SOTA results on IAM-Test set when training data is limited, also Easter2.0 has very small number of trainable parameters compared to other solutions. The proposed architecture can be used in search of smaller, faster and efficient OCR/HTR solutions when available annotated data is limited.

The following steps can help setting up Easter2 fast:

• Clone the repository
• Install requirements as per the file requirements.txt
• Download checkpoint from release, and put it inside /weigths directory (if you want to finetune the older checkpoint or make it False in the config.py)
• Download you dataset in the data folder
• Use the custom_dataset.ipynb to convert your dataset into the required format. CHnage the path inside it will make a file with all the paths of the files. (./img_path.png words_name)
• Modify /src/config.py as per your needs
• Using the created file in the previous step, and add its path in the /src/data_loader.py file
• Run the train() function from /src/easter_model.py or just simply execute the easter_model.py
• Sample training and testing notebooks are given in /notebooks directory

import tensorflow as tf
print("Num GPUs Available: ", len(tf.config.list_physical_devices('GPU')))

I have used the work of chaudhary2022easter2. If you find our work helpful, please cite the following:

@article{chaudhary2022easter2,
  title={Easter2. 0: Improving convolutional models for handwritten text recognition},
  author={Chaudhary, Kartik and Bali, Raghav},
  journal={arXiv preprint arXiv:2205.14879},
  year={2022}
}