PaddleOCR icon indicating copy to clipboard operation
PaddleOCR copied to clipboard

Awesome multilingual OCR toolkits based on PaddlePaddle (practical ultra lightweight OCR system, support 80+ languages recognition, provide data annotation and synthesis tools, support training and de...

Results 1088 PaddleOCR issues
Sort by recently updated
recently updated
newest added

使用原始的cml配置文件会报错: 修改了以下几处地方,可正常运行 1.PPLCNetNew改成PPLCNetV3 2. 添加一行det: true (两个stu网络都需要加) 3. cal_metric_during_train 由true改为fasle

# 补充Satrn识别模型TIPC ## 任务描述 ### 任务背景 - 新增的Satrn模型缺少TIPC,完成tipc有利于上手训推全流程自动化脚本验证过程 ### 完成步骤 1. 参照[https://github.com/PaddlePaddle/PaddleOCR/pull/10843](https://github.com/PaddlePaddle/PaddleOCR/pull/10843)增加TIPC中对应的内容,并且可以跑通[test_train_inference_python.sh](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/test_tipc/test_train_inference_python.sh)。 ### 提交内容: 1. 提交类似示例中的修改,实现Satrn模型的TIPC。

请问大家有没有遇到和我一样的问题?如何解决 数据集是XFUND,ser_vi_layoutxlm_xfund_zh.yml是官方的。 环境如下: win10, cuda 11.6 cundd 8.4.0.27 ![image](https://github.com/PaddlePaddle/PaddleOCR/assets/56833624/539e52a8-a48d-4989-b225-988e2f8470ad) 之前有试过:paddlepaddle-gpu==2.3.2,paddlenlp-2.3.0.dev0,也出现同样的报错, 我的全部日志如下: D:\DFSJ\OCR\project\paddle-ocr-gpu>python tools/train.py -c configs/kie/vi_layoutxlm/ser_vi_layoutxlm_xfund_zh.yml [2023/10/25 18:25:53] ppocr INFO: Architecture : [2023/10/25 18:25:53] ppocr INFO: Backbone : [2023/10/25 18:25:53] ppocr...

Currently, the `BASE_DIR` in the code is hardcoded as `os.path.expanduser("~/.paddleocr/")`. It would be beneficial to allow users to specify the `BASE_DIR` through an environmental variable, giving them more flexibility in...

请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem - 系统环境/System Environment:CPU 12th Gen Intel(R) Core(TM) i7-12700H GPU Nvidia 3060 6G - 版本号/Version:Paddle:2.4.2 PaddleOCR:2.6.2 问题相关组件/Related components:PaddleOCR C++部署 - 运行指令/Command...

Hi all, I have a large amount of images to process, is there a way to do inference using a batch and maybe a dataloader? Classifying one image at a...

#### 问题描述 当前CI流程逻辑下,任何文件改动的PR都会出发单元检测和格式检测。这有些多余了。 如果我只改动文档类文件,没有必要跑一遍单元测试和格式检测了。 #### 解决方案 在Action配置文件,添加后缀过滤规则,排除md文件,md文件改动时,不触发CI。

#### 问题描述 / Problem Description 下载最新发布的表格识别模型进行推理,然后按照文档来运行会报错 #### 运行环境 / Runtime Environment - OS: - Paddle: 2.3.2.post112 - PaddleOCR: 2.7.0.2 #### 复现代码 / Reproduction Code python table/predict_table.py --det_model_dir=../weights/ch_PP-OCRv3_det_infer --rec_model_dir=../weights/ch_PP-OCRv3_rec_infer --table_model_dir=../weights/ch_ppstructure_openatom_SLANetv2_infer --rec_char_dict_path=../ppocr/utils/ppocr_keys_v1.txt...

#### 问题描述 / Problem Description windows 本地不报错,linux报错 ValueError: A global iterator flag was passed as a per-operand flag to the iterator constructor #### 运行环境 / Runtime Environment - OS: linux...

#### 问题描述 / Problem Description ``` InvalidArgumentError: Broadcast dimension mismatch. Operands could not be broadcast together with the shape of X = [6, 96, 3, 12] and the shape of...