apsis-net

Apsis-net is a Bengali language ocr system for Printed Documents developed at Apsis Solutions limited

The full system is focused on bengali text recognition only

• Text recognition:
  • Bangla Text : ApsisNet
    • ApsisNet is a model developed at Apsis Solutions Limited.
    • It is used by bbOCR as the recognition model
    • ApsisNet is found to be the best among other available recognition models (such as tesseract and easyOCR) in the linked paper

Installation

As module/pypi package

cpu installation

pip install apsisnet

gpu installation

It is recommended to use conda environment . Specially for GPU.

• installing cudatoolkit and cudnn:

conda install cudatoolkit
conda install cudnn

• installing packages

pip install apsisnet

• exporting environment variables

mkdir -p $CONDA_PREFIX/etc/conda/activate.d
echo 'export LD_LIBRARY_PATH=$CUDNN_PATH/lib:$CONDA_PREFIX/lib/:$LD_LIBRARY_PATH' >> $CONDA_PREFIX/etc/conda/activate.d/env_vars.sh

Building from source : Linux/Ubuntu

It is recommended to use conda environment .

• clone the repository :

git clone https://github.com/mnansary/apsisnet.git
cd apsisnet

• create a conda environment:

conda create -n apsisnet python=3.9

• activate conda environment:

conda activate apsisnet

• cpu installation :

bash install.sh cpu

• gpu installation :

bash install.sh gpu

Useage

Apsisnet : Bangla Recognizer

• useage

from apsisnet import ApsisNet
bnocr=ApsisNet()
bnocr.infer(crops)

• docstring for ApsisNet.infer

"""
Perform inference on image crops.

Args:
    crops (list[np.ndarray]): List of image crops.
    batch_size (int): Batch size for inference (default: 32).
    normalize_unicode (bool): Flag to normalize unicode (default: True).

Returns:
    list[str]: List of inferred texts.
"""

check useage/useage.ipynb for examples

TESTED GPU INFERENCE SERVER CONFIG

OS          : Ubuntu 20.04.6 LTS      
Memory      : 62.4 GiB 
Processor   : Intel® Xeon(R) Silver 4214R CPU @ 2.40GHz × 24    
Graphics    : NVIDIA RTX A6000/PCIe/SSE2
Gnome       : 3.36.8

License

Contents of this repository are restricted to non-commercial research purposes only under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0).

25-04-2024

• changed onnxruntime-gpu version