PaddleX/latest/module_usage/tutorials/ocr_modules/table_structure_recognition

[giscus[bot]]

PaddleX/latest/module_usage/tutorials/ocr_modules/table_structure_recognition

https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/ocr_modules/table_structure_recognition.html

[Danee-wawawa]

您好，非常感谢您的工作，我有几个问题想问下：想问下上述模型列表中提供的SLANet表格识别模型和paddleocr提供的en_ppstructure_mobile_v2.0_SLANet和ch_ppstructure_mobile_v2.0_SLANet是什么对应关系嘛～

[cuicheng01]

SLANet对应ch_ppstructure_mobile_v2.0_SLANet，后续paddleocr和paddlex的名字命名会进行一次大的统一，统一使用paddlex的命名方式

[SunboZ]

训练的时候，main.py是哪个

[cuicheng01]

需要clone仓库，这块在文档中有描述，可以查看安装文档

[SunboZ]

好的，非常感谢

[useOnePunchMan]

我的表里有很多的合并单元格，使用SLANeXt_wired及SLANeXt_wireless推理模型识别的bbox是一样的，绘制出来效果很差，但为什么生成的html又几乎与原来表中的格式几乎相近？

[ousandada]

您好请问这个问题你解决了吗？我的没有合并表识别的也很差，save_to_img结果很不好，

[garrettXu]

from paddlex import create_model
model = create_model(model_name="SLANeXt_wired")
output = model.predict(input="abc.jpg")
for res in output:
res.print(json_format=False)
res.save_to_json("./output/res.json") 这样使用，报错了。 warnings.warn(warning_message)
WARNING: OMP_NUM_THREADS set to 12, not 1. The computation speed will not be optimized if you use data parallel. It will fail if this PaddlePaddle binary is compiled with OpenBlas since OpenBlas does not support multi-threads.
PLEASE USE OMP_NUM_THREADS WISELY.
Traceback (most recent call last):
File "/root/autodl-tmp/paddlx_demo.py", line 2, in
model = create_model(model_name="SLANeXt_wired")
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/miniconda3/lib/python3.12/site-packages/paddlex/model.py", line 28, in create_model
return _ModelBasedInference(
^^^^^^^^^^^^^^^^^^^^^
File "/root/miniconda3/lib/python3.12/site-packages/paddlex/model.py", line 58, in init
self._predictor = create_predictor(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/miniconda3/lib/python3.12/site-packages/paddlex/inference/models/init.py", line 77, in create_predictor
return BasePredictor.get(model_name)(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/miniconda3/lib/python3.12/site-packages/paddlex/inference/models/table_structure_recognition/predictor.py", line 37, in init
self.preprocessors, self.infer, self.postprocessors = self._build()
^^^^^^^^^^^^^
File "/root/miniconda3/lib/python3.12/site-packages/paddlex/inference/models/table_structure_recognition/predictor.py", line 56, in _build
infer = self.create_static_infer()
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/miniconda3/lib/python3.12/site-packages/paddlex/inference/models/base/predictor/base_predictor.py", line 248, in create_static_infer
return PaddleInfer(self.model_dir, self.MODEL_FILE_PREFIX, self._pp_option)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/miniconda3/lib/python3.12/site-packages/paddlex/inference/models/common/static_infer.py", line 319, in init
self.predictor = self._create()
^^^^^^^^^^^^^^
File "/root/miniconda3/lib/python3.12/site-packages/paddlex/inference/models/common/static_infer.py", line 486, in _create
predictor = paddle.inference.create_predictor(config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ValueError: (InvalidArgument) The size of inputs must be equal to 1.
[Hint: Expected input_size == 1, but received input_size:2 != 1:1.] (at paddle/fluid/pir/dialect/operator/ir/pd_op3.cc:16510)

[dolphincats]

表格结构识别模型无法转为OM，和其他模型配合om和onnxruntime混合推理时候，冲突报错？