frotms
frotms
构建好网络加载好权重后，用torch.onnx.export就可以转了
按文档能推理起来，在推理代码后面加这一句就可以了
模型是paddleocr训练好的，训练模型对应的数据集可以从paddleocr项目查看
@quliulangle 请提供原始图像以及两者结果可视化图像看下
@songchuanle-1 模型转换使用的是paddle的训练模型，不是推理模型
https://github.com/frotms/PaddleOCR2Pytorch/blob/main/doc/doc_ch/inference.md#%E5%85%B6%E4%BB%96%E8%AF%86%E5%88%AB%E6%A8%A1%E5%9E%8B
用paddleocr训练即可
@xudongwang0828 I take it as a fisheye distortion problem. 3D modeling with a virtual camera and fisheye camera model may rectify curve text in image. For details, see the Reference...
> > @xudongwang0828 I take it as a fisheye distortion problem. 3D modeling with a virtual camera and fisheye camera model may rectify curve text in image. For details, see...
> > > @xudongwang0828 I take it as a fisheye distortion problem. 3D modeling with a virtual camera and fisheye camera model may rectify curve text in image. For details,...