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't we change input data in configs file? #8

Open
ghost opened this issue May 17, 2020 · 4 comments
Open

Can't we change input data in configs file? #8

ghost opened this issue May 17, 2020 · 4 comments

Comments

@ghost
Copy link

ghost commented May 17, 2020

We have:

# Input data
input_w: 1024
input_h: 1024

in configs/001.yml file. Can't we change it to other pixel formats, for example, 3780x6720 or retaining the original size of images? Please feel free to throw some light on why it can't accept the image as is.

I tried changing it to given example but the program was immediately killed.

It gives weird outputs when the size is changed to 1024x1024. Check the images. The first one is original, second one is the output by the model.

Screenshot from 2020-05-17 18-21-24
Screenshot from 2020-05-17 18-03-02

@Xu-Yao
Copy link
Contributor

Xu-Yao commented May 24, 2020

Hi, sorry for my late reply. This model has only viewed aligned face images of 1024 resolution at training time. We train a separate model for each resolution (256, 512 and 1024). The faces are aligned and cropped using dlib. The true resolution of the faces in the images that you provided is about 256x256. So I processed them with our model at 256 resolution. Here are the results (aged to 65):
001001_age_65
002002_age_65

@luoqiuluoqiu
Copy link

Hi, sorry for my late reply. This model has only viewed aligned face images of 1024 resolution at training time. We train a separate model for each resolution (256, 512 and 1024). The faces are aligned and cropped using dlib. The true resolution of the faces in the images that you provided is about 256x256. So I processed them with our model at 256 resolution. Here are the results (aged to 65):
001001_age_65
002002_age_65

Hello, can you provide the face correction code and
the models parameters with 256 and 512 resolutions ?

@ghost
Copy link
Author

ghost commented Jun 5, 2020

Hi, sorry for my late reply. This model has only viewed aligned face images of 1024 resolution at training time. We train a separate model for each resolution (256, 512 and 1024). The faces are aligned and cropped using dlib. The true resolution of the faces in the images that you provided is about 256x256. So I processed them with our model at 256 resolution. Here are the results (aged to 65):
001001_age_65
002002_age_65

Thank you for the reply @Xu-Yao . Can you provide the face correction code as @luoqiuluoqiu pointed out?

@kishanios123
Copy link

Hello there,

i have same issue in smaller size ... changed width height in config to 256... and changed in image file also...

and getting output not proper ... can any one help me please ...
Screenshot 2020-08-08 at 12 50 12 PM

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