-
PP-OCRv3识别阶段的resize_norm_img_svtr函数采用的resize方式会使图片的宽高比发生变化,图片发生形变,导致识别出现问题啊,改成保持宽高比不变的形式就识别对了
改之前:
改之后:
-
i trained svtrnet for bangla text recognition task as discussed here : https://github.com/PaddlePaddle/PaddleOCR/issues/6632 ,i trained the model for long time(6-7 days) and then got acc 0.8.
after g…
-
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 Cod…
-
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:windows11
- 版本号/Version:Paddle: PaddleOCR: 问题相关组件/Related components:2.6.0
- 运行…
-
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:win10 cuda10.2 cudnn7.6.5 gtx1050ti
- 版本号/Version:Paddle:gpu-2.3.0 PaddleOCR:2.…
-
按照教程更新PaddleOCR V3版本whl包:
paddleocr 2.5.0.3
paddlepaddle 2.3.0
paddlepaddle-gpu 2.3.0
配置参数:
Namespace(alpha=1.0, benchmark=False, b…
-
根据quickstart配置环境后,运行 paddleocr --image_dir ./1_cut.bmp --lang=en
之后程序一直卡在一个地方也没报错
`[2022/06/21 20:54:11] ppocr DEBUG: Namespace(help='==SUPPRESS==', use_gpu=True, ir_optim=True, use_tensorrt=False, …
-
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:ubuntu18.04 cuda 10.2 cudnn8 python3.7 tensorrt 7.2.3.4
- paddle2onnx 0…
-
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:macOS 12.4 (21F79)
- 版本号/Version:Paddle:2.3.1 PaddleOCR: 41ccb1ff 问题相关组件/Related…
-
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:RTX A5000 + ubuntu 18.04 + cuda 11.2
- 版本号/Version:Paddle: PaddleOCR: 问题相关组件/Re…