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

Train my own data #63

Open
mauione1 opened this issue Apr 10, 2018 · 5 comments
Open

Train my own data #63

mauione1 opened this issue Apr 10, 2018 · 5 comments

Comments

@mauione1
Copy link

Hi,
I am new to MobileNet. Could some one help me to train my own data for caffe. I dont see any training script here.

@qiuxiaqing
Copy link

I also want to know how to generate the new caffemodle with my own database. Have you solved this problem? if yes, would you like to give me some suggestion? Thank you a lot

@HilmiK
Copy link

HilmiK commented May 9, 2018

You should modify these deploy.prototxt file to create train.prototxt where some examples are given in original caffe repository such as alexnet, vgg16. You can also convert your data into lmdb file.

@qiuxiaqing
Copy link

yes thank you. you are right! now, I have do this ,but i want to know what's your mean_value in train.prototxt. (shicai's mean_value: [103.94, 116.78, 123.68]......then, must I compute my data's mean_value in this part???) thank you
can you tell me how to compute the mean_value? Thank you so much

@HilmiK
Copy link

HilmiK commented May 10, 2018

You can leave them as they are. They should be computed from Imagenet and should not matter a lot to change them. U can even comment them, the network should still work.

@qiuxiaqing
Copy link

OK, That may not be what I want . Thank you all the same

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

3 participants