Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

自己训练的MTCNN检测模型,ncnn版本与caffe版本差异很大 #198

Closed
zhiweizhong opened this issue Nov 17, 2017 · 4 comments
Closed

Comments

@zhiweizhong
Copy link

zhiweizhong commented Nov 17, 2017

RT
看到之前的Issue中说原始的MTCNN模型是col-major的,我现在是用caffe训练的模型(在MTCNN模型上fine-tune),按理来说应该是row-major的,不需要转换了吧。。
我是训练mtcnn检测手势,用Rnet单独测试一张图片得分很低,
`const float mean_vals_[3] = {127.5f, 127.5f, 127.5f};
const float norm_vals_[3] = {0.0078125f, 0.0078125f, 0.0078125f};
img.substract_mean_normalize(mean_vals_, norm_vals_);
ncnn::Mat in;
resize_bilinear(img, in, 24, 24);
ncnn::Extractor ex = Rnet_.create_extractor();
ex.set_light_mode(true);
ex.input("data", in);
ncnn::Mat confidence, regression;

ex.extract("prob1", confidence);
std::cout << "rnet score: " << *(confidence.data + confidence.cstep) << std::endl;`

@nihui
我的模型文件如下:

Rnet_model.zip

@rivercn
Copy link

rivercn commented Dec 3, 2017

你的模型训练环境是什么?

@nihui
Copy link
Member

nihui commented Jan 4, 2018

@wolf345
Copy link

wolf345 commented Feb 2, 2018

@zhiweizhong 请问一下,你最后问题解决了吗?怎么解决的

@nihui
Copy link
Member

nihui commented Dec 18, 2018

mtcnn 可以参考 https://github.com/moli232777144/mtcnn_ncnn

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants