Views
No views yet
| Model | Recognition Avg Accuracy(%) | Model Storage Size (M) | Introduction |
|---|---|---|---|
| en_PP-OCRv3_mobile_rec | 70.69 | 7.8 M | An ultra-lightweight English recognition model trained based on the PP-OCRv3 recognition model, supporting English and numeric character recognition. |
1# for CUDA11.8
2python -m pip install paddlepaddle-gpu==3.0.0 -i https://www.paddlepaddle.org.cn/packages/stable/cu118/
3
4# for CUDA12.6
5python -m pip install paddlepaddle-gpu==3.0.0 -i https://www.paddlepaddle.org.cn/packages/stable/cu126/
6
7# for CPU
8python -m pip install paddlepaddle==3.0.0 -i https://www.paddlepaddle.org.cn/packages/stable/cpu/python -m pip install paddleocr1paddleocr text_recognition \
2 --model_name en_PP-OCRv3_mobile_rec \
3 -i https://cdn-uploads.huggingface.co/production/uploads/681c1ecd9539bdde5ae1733c/QmaPtftqwOgCtx0AIvU2z.png1from paddleocr import TextRecognition
2model = TextRecognition(model_name="en_PP-OCRv3_mobile_rec")
3output = model.predict(input="QmaPtftqwOgCtx0AIvU2z.png", batch_size=1)
4for res in output:
5 res.print()
6 res.save_to_img(save_path="./output/")
7 res.save_to_json(save_path="./output/res.json"){'res': {'input_path': '/root/.paddlex/predict_input/QmaPtftqwOgCtx0AIvU2z.png', 'page_index': None, 'rec_text': 'the number of model parameters and FLOPs get larger, it', 'rec_score': 0.990352988243103}}
1paddleocr ocr -i https://cdn-uploads.huggingface.co/production/uploads/681c1ecd9539bdde5ae1733c/c3hSldnYVQXp48T5V0Ze4.png \
2 --text_recognition_model_name en_PP-OCRv3_mobile_rec \
3 --use_doc_orientation_classify False \
4 --use_doc_unwarping False \
5 --use_textline_orientation True \
6 --save_path ./output \
7 --device gpu:0 1{'res': {'input_path': '/root/.paddlex/predict_input/c3hSldnYVQXp48T5V0Ze4.png', 'page_index': None, 'model_settings': {'use_doc_preprocessor': True, 'use_textline_orientation': True}, 'doc_preprocessor_res': {'input_path': None, 'page_index': None, 'model_settings': {'use_doc_orientation_classify': False, 'use_doc_unwarping': False}, 'angle': -1}, 'dt_polys': array([[[252, 172],
2 ...,
3 [254, 241]],
4
5 ...,
6
7 [[665, 566],
8 ...,
9 [663, 601]]], dtype=int16), 'text_det_params': {'limit_side_len': 64, 'limit_type': 'min', 'thresh': 0.3, 'max_side_limit': 4000, 'box_thresh': 0.6, 'unclip_ratio': 1.5}, 'text_type': 'general', 'textline_orientation_angles': array([0, ..., 0]), 'text_rec_score_thresh': 0.0, 'rec_texts': ['The moon tells the sky', 'The sky tells the sea', 'The sea tells the tide', 'And the tide tells me', 'Lemn Sissay'], 'rec_scores': array([0.99890286, ..., 0.99840254]), 'rec_polys': array([[[252, 172],
10 ...,
11 [254, 241]],
12
13 ...,
14
15 [[665, 566],
16 ...,
17 [663, 601]]], dtype=int16), 'rec_boxes': array([[252, ..., 241],
18 ...,
19 [663, ..., 612]], dtype=int16)}}save_path. The visualization output is shown below:
1from paddleocr import PaddleOCR
2
3ocr = PaddleOCR(
4 text_recognition_model_name="en_PP-OCRv3_mobile_rec",
5 use_doc_orientation_classify=False, # Use use_doc_orientation_classify to enable/disable document orientation classification model
6 use_doc_unwarping=False, # Use use_doc_unwarping to enable/disable document unwarping module
7 use_textline_orientation=True, # Use use_textline_orientation to enable/disable textline orientation classification model
8 device="gpu:0", # Use device to specify GPU for model inference
9)
10result = ocr.predict("https://cdn-uploads.huggingface.co/production/uploads/681c1ecd9539bdde5ae1733c/c3hSldnYVQXp48T5V0Ze4.png")
11for res in result:
12 res.print()
13 res.save_to_img("output")
14 res.save_to_json("output")PP-OCRv5_server_rec, so it is needed that specifing to en_PP-OCRv3_mobile_rec by argument text_recognition_model_name. And you can also use the local model file by argument text_recognition_model_dir. For details about usage command and descriptions of parameters, please refer to the Document.