PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/seal_recognition #3882
Replies: 5 comments 5 replies
-
|
使用“印章文本识别产线”无法替换模型 我在使用"印章文本识别产线"时，通过 paddlex --get_pipeline_config seal_recognition --save_path ./my_path 获取了配置文件，由于测试过程中文本识别模块准确率不达预期，所以我想通过替换文本识别模型为 PP-OCRv4_server_rec_doc来测试准确率是否会提升。修改后的配置文件： Global: Pipeline: 但是我看PP-OCRv4_server_rec_doc这个模型应该是seal_recognition这条产线所支持的，是否哪里配置不对？ |
Beta Was this translation helpful? Give feedback.
-
|
请问您自己的文本识别模型，使用多大的数据集来训练的啊？效果怎么样啊 |
Beta Was this translation helpful? Give feedback.
-
|
官方提供的示例，好像只使用了版面检测模块RT-DETR-H_layout_3cls，印章文本检测模块PP-OCRv4_server_seal_det，文本识别模块PP-OCRv4_server_rec。 文本图像矫正模块，文档图像分类这些模块都没有使用。 |
Beta Was this translation helpful? Give feedback.
-
|
请问怎么重新加载修改后的产线配置文件？ |
Beta Was this translation helpful? Give feedback.
-
|
哦哦看到了解决了 |
Beta Was this translation helpful? Give feedback.
-
|
请问下 通用OCR产线 高性能服务化部署 ，推荐的 服务器配置 是什么？最低配置是什么 ？ 感谢 |
Beta Was this translation helpful? Give feedback.
-
|
这个没办法推荐，因为确实不知道最低的配置应该是什么，大多数的GPU都是可以的 |
Beta Was this translation helpful? Give feedback.
-
|
请问可以修改seal类别为其他类别吗 |
Beta Was this translation helpful? Give feedback.
-
|
请问50系显卡预计什么时候可以适配呢 |
Beta Was this translation helpful? Give feedback.
-
|
数据量很大，不适合下载到本地的话，如何改写Dataset类进行微调呢 |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/seal_recognition
https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/seal_recognition.html
Beta Was this translation helpful? Give feedback.
All reactions