Felix92/doctr-dummy-tf-parseq

1 2---3language: en4---5 6<p align="center">7<img src="https://doctr-static.mindee.com/models?id=v0.3.1/Logo_doctr.gif&src=0" width="60%">8</p>9 10**Optical Character Recognition made seamless & accessible to anyone, powered by TensorFlow 2 & PyTorch**11 12## Task: recognition13 14https://github.com/mindee/doctr15 16### Example usage:17 18```python19>>> from doctr.io import DocumentFile20>>> from doctr.models import ocr_predictor, from_hub21 22>>> img = DocumentFile.from_images(['<image_path>'])23>>> # Load your model from the hub24>>> model = from_hub('mindee/my-model')25 26>>> # Pass it to the predictor27>>> # If your model is a recognition model:28>>> predictor = ocr_predictor(det_arch='db_mobilenet_v3_large',29>>>                           reco_arch=model,30>>>                           pretrained=True)31 32>>> # If your model is a detection model:33>>> predictor = ocr_predictor(det_arch=model,34>>>                           reco_arch='crnn_mobilenet_v3_small',35>>>                           pretrained=True)36 37>>> # Get your predictions38>>> res = predictor(img)39```40