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

How can I train custom datasets? #9

Open
blackCmd opened this issue Apr 21, 2021 · 2 comments
Open

How can I train custom datasets? #9

blackCmd opened this issue Apr 21, 2021 · 2 comments

Comments

@blackCmd
Copy link

Thanks for your repo.
It's amazing work.

I want to train custom dataset.
How can I do?

@shashank-s-adsule
Copy link

hi i wanted to train this model on an custom dataset of 2835 image [FDDB dataset] for Face Detection.

how can i do so also i have some querys:

1. how create custom dataset .pkl file for this model
2. when analyzed the **wiferface.pkl** file provided on the README the meta data i got:
   1. data -> dict()
      1. keys: index [0...56922]
      2. values: -> list 
         1.  pos 0: image_buffer (N*1 array-> shape=>(vary,1))
         2.  pos 1: flag [0: -ve image, 1: +ve image]
         3.  pos 2: bbox [N*4 => number of bound  boxes] 
    so i want to know how to create image_buffer if i have got an image of np.array() from cv2

if possible please guide me through this part

@shashank-s-adsule
Copy link

hi i wanted to train this model on an custom dataset of 2835 image [FDDB dataset] for Face Detection.

how can i do so also i have some querys:

1. how create custom dataset .pkl file for this model
2. when analyzed the **wiferface.pkl** file provided on the README the meta data i got:
   1. data -> dict()
      1. keys: index [0...56922]
      2. values: -> list 
         1.  pos 0: image_buffer (N*1 array-> shape=>(vary,1))
         2.  pos 1: flag [0: -ve image, 1: +ve image]
         3.  pos 2: bbox [N*4 => number of bound  boxes] 
    so i want to know how to create image_buffer if i have got an image of np.array() from cv2

if possible please guide me through this part

the image buffer is create using cv2.imencode() function.
after that the encodeed_img is reshape using:
encode_img=encode_img.reshape(-1,1)

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