we0091234 / crnn_plate_recognition Goto Github PK
View Code? Open in Web Editor NEWcrnn chinese_plate_recognition
crnn chinese_plate_recognition
如题,请教一下如何获取识别后的车牌每个字符跟颜色的置信度
做好数据处理等前置工作后,执行训练脚本 python train.py --cfg OWN_config.yaml 提示模型尺寸不正确,示例代码train.py 中构建的三种模型都不正确,如何约束这个尺寸问题?
下面展示各个模型的错误:
1 -- build_lprnet
model = build_lprnet(num_classes=len(plate_chr))
错误为:RuntimeError: Given groups=1, weight of size [64, 3, 3, 3], expected input[32, 1, 32, 160] to have 3 channels, but got 1 channels instead
2 -- myNet_ocr
model = myNet_ocr(num_classes=len(plate_chr),cfg=cfg)
错误为:RuntimeError: Given groups=1, weight of size [32, 3, 5, 5], expected input[32, 1, 32, 160] to have 3 channels, but got 1 channels instead
3 -- crnn
model = crnn.get_crnn(config,cfg=cfg)
错误为:RuntimeError: Given groups=1, weight of size [32, 3, 5, 5], expected input[32, 1, 32, 160] to have 3 channels, but got 1 channels instead
大家是否遇到同样的问题,有解决方案吗?
~/Documents/GitHub/crnn_plate_recognition/ [master] python train.py --cfg lib/config/360CC_config.yaml
Traceback (most recent call last):
File "train.py", line 8, in <module>
import lib.models.crnn as crnn
ImportError: No module named lib.models.crnn
您好,请问训练过程中使用的myNet_ocr模型与CRNN模型有联系吗?怎么替换为CRNN模型呢?感谢大佬。
您好,我在训练过程中,saved_model/best.pt 一直没有更新,另外output的checkpoint 文件夹无法进入。请问有哪些问题需要注意的?
请问博主,是否还有别的训练方法啊,我这边按照您提供的readme训练,虽然模型精度也能达到98%,但是模型的泛化性还是比您训练的模型差一点点,特别是双层车牌效果没您的好,请问您也是用的您所提供的dataset训练的吗?还是说加过其它额外数据?或者是在调参上面有一些技巧呢?谢谢。
请教大佬,pt导出onnx, 输出结果是 [1,21,78]与tensorrt的导出engine,进行推理结果不对
python export.py --weights saved_model/plate_rec.pth --save_path saved_model/plate_rec.onnx --simplify
请问 large模型有放出嘛?
请问颜色数据集训练的时候,数据集txt文件怎么生成的啊?
我很想用此代码来训练文本识别,文本的长度是不固定的。但是 我们训练和 ncnn中的识别代码,都是宽度是固定的168,对于车牌这种还可以。要是 文本变长的,我们应该怎么办啊,求教。
train.py parser.add_argument('--img_w',type=int,default=168,help='width')
脚本
OW: 280 # origial width: 280
H: 32
W: 100 # resized width: 160
Hi @we0091234,
I'm Arunkumar from India. I can't connect with you on both WeChat and QQ group. Are there any other ways to connect with you? I have some doubts I need to discuss.
Thanks!
Is there any reference to someone's knowledge which explains how to compute the accuracy of the LPRNet model output? I'm not interested into computing WER or CER, but an overall accuracy value for the LPRNet output
大佬好.
用来训练车牌识别的数据事先需要仿射变换处理吗
老哥,这个算法可以识别双层的车牌吗?具体是怎么做的?
kpt-label 为2 报错,这个是kpt写死了只能为4么
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.