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

Can i use my own new images to train the model? #31

Closed
shashi19081994 opened this issue Nov 4, 2017 · 1 comment
Closed

Can i use my own new images to train the model? #31

shashi19081994 opened this issue Nov 4, 2017 · 1 comment

Comments

@shashi19081994
Copy link

Can i use my own new images to train the model? I am working on cnn head pose estimation. your work is excellent. I was stuck with a question that "Can i use my own new images to train the model? " whether this deepgaze will help me so in doing? Thank you.

@mpatacchiola
Copy link
Owner

Dear @shashi19081994

At the moment there is not a file that you can use to train the model. However there is an implementation of a trainer for aDeep MLP here in the example folder. You can slightly modify this example including the model and the weights used in the Deepgaze CNN head_pose_estimation.py package. If you have familiarity with Tensorflow it should be straightforward.

Once you have the file with the correct model and weights then it is necessary to pre-process your images. In the same example folder you can find some parser used on the Prima and AFLW datasets. You should modify one of the parser in order to resize and crop your images and then store them in a pickle file, which is the format used by the trainer.

At the end of the training you need to save the Tensorflow meta-files which can be loaded in Deepgaze or in your personal model.

Best, Massimiliano

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

2 participants