-
Hi @valeman
I've been reading through your codebase and I figured I will need your clarifications on some part of it, I'll be using this thread to post my questions.
Also, I haven't _**really**…
-
When trying to print with an LBP2900 from Raspbian 10 armhf, the printer gets stuck in a loop. Other versions of Raspbian may be affected as well.
CUPS debug logs reveal that the loop is a status c…
-
您好,我用yolov4 pytorch训练识别车牌,标签只有一类,但是训练后只能检测一个物体,不能检测多个,训练集采用的是CCPD,每张图片内也只有一个物体。请问您知道是什么原因吗
-
[基于Python预测引擎推理
-> 1.超轻量中文检测模型推理教程](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.0/doc/doc_ch/inference.md#1-%E8%B6%85%E8%BD%BB%E9%87%8F%E4%B8%AD%E6%96%87%E6%A3%80%E6%B5%8B%E6%A8%A1%E5%…
-
这是我自己加了"港", "学", "使", "警", "澳", "挂", "军", "北", "南", "广", "沈", "兰", "成", "济", "海", "民", "航", "空", "应", "急", "试", "超"这些字符后训练后测试的结果,我使用的是100多万张真实的数据训练的,结果惨不忍睹,我搞不清楚是什么原因。
Successful to build network!
…
rawli updated
3 years ago
-
自己训练的mv3_db_det模型,利用原始模型的检测效果可以接受,然而推理后得到的模型测试效果差距很大
这是什么原因呢?怎么解决?
原始模型:![0160](https://user-images.githubusercontent.com/23338843/84744096-5f550380-afe5-11ea-96b5-198263aa3536.jpg)
推理模型:
![det_re…
-
https://github.com/PaddlePaddle/PaddleOCR/blob/fa32b9b184528d446a3aea14a750a3912491dfd5/ppocr/utils/utility.py#L55
https://github.com/PaddlePaddle/PaddleOCR/blob/fa32b9b184528d446a3aea14a750a391249…
-
您好,一直关注您的算法,这个车牌关键点检测的训练样本可否分享一下。
-
使用rec_chinese_lite_train_v2.0.yml训练识别模型,对车牌字符进行识别,标准车牌比例是140*440(宽高比3.14)。
我基于CCPD数据集截取车牌区域后,设置RecResizeImg.image_shape: [3, 32, 320]或[3, 32, 100]时可训练,但这个尺寸和车牌相差比较远,重新设置.yml文件,按车牌比例的大小,设置为 [3, 70, 22…
-
How to process the dataset CCPD to get the dataset.json?