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

python3 tools/train.py -c configs/picodet/legacy_model/application/layout_analysis/picodet_lcnet_x1_0_layout.yml --eval - 系统环境/System Environment:python 3.7.13 Linux version 5.15.0-52-generic (buildd@lcy02-amd64-045) (gcc (Ubuntu 9.4.0-1ubuntu120.04.1) 9.4.0, GNU ld (GNU Binutils for Ubuntu) 2.34) #5820.04.1-Ubuntu SMP Thu Oct 13 13:09:46...

1. 配置文件是ch_PP-OCRv3_rec_distillation.yml; 2. 自己的训练集和验证集是12000张表格线切分后的图片,宽高比大于10; 3. 预训练模型来自[https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_train.tar](url); 4. 看训练日志,准确率达到0.86,但用tools/infer_rec.py脚本测试训练出的模型,效果比较差; 5. 自己是按照[https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.6/doc/doc_ch/recognition.md#21-%E5%90%AF%E5%8A%A8%E8%AE%AD%E7%BB%83](url)来做的finetune,感觉应该没问题啊 6. PP-OCRv3的推理模型,效果很好

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: - 版本号/Version:Paddle: PaddleOCR: 问题相关组件/Related components: - 运行指令/Command Code: - 完整报错/Complete Error Message: contours, _ = cv2.findContours((bitmap *...

detection

针对误用参数及早报错

contributor

你好,请问一下,PPOCRV3识别训练过程中,没有出现关于Loss、Recall等数据,这一块是删除了吗? ![image](https://user-images.githubusercontent.com/87969070/201275830-6b712ee2-97d6-4f16-911b-fe304b50f531.png)

**安装命令:** ```pip3 install -r requirerents.txt 尝试使用了以下方法: * pip3 install scikit-image * 下载scikit-image包,然后再安装 * 更新pip * 换源 ![656c64cbf6a681f49b8c29a9179d5df](https://user-images.githubusercontent.com/117959671/201248894-e5882058-f8a8-4789-9ac0-8b521941b482.jpg)

想问下有针对中文论文的版面分析预训练模型吗,PubLayNet/ppyolov2_r50vd_dcn_365e_publaynet/config 这个是英文的,对中文效果不是太好

ImportError: cannot import name 'to_excel' from 'paddleocr' (f:\software\python37\lib\site-packages\paddleocr\__init__.py)

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 Hi, How can i filter a folder with images with only images that contain text. So keep only images with...

- 系统环境/System Environment:nvidia jetson - 版本号/Version: Paddle:2.3 gpu PaddleOCR:2.6 - 运行指令/Command Code: -` python3.6 tools/infer/predict_rec.py --image_dir='test/images/0' --rec_model_dir='./rec_visionlan/' --rec_algorithm='VisionLAN' --rec_image_shape='3,64,256' --rec_char_dict_path='./ppocr/utils/en_dict.txt' --use_space_char=False --use_gpu=False` - 完整报错/Complete Error Message: `ppocr INFO: Traceback (most...