TencentARC/GFPGAN
1
0
Fork 0
mirror of https://github.com/TencentARC/GFPGAN.git synced 2026-04-27 00:01:00 +00:00
Code Issues 476 Packages Projects Releases 8 Wiki Activity

Restore face without colorizing them #14

New Issue
Closed
opened 2026-01-29 21:36:37 +00:00 by claunia · 18 comments
claunia commented 2026-01-29 21:36:37 +00:00
Owner
Copy Link

Originally created by @syfbme on GitHub (Jul 5, 2021).

I wonder if your model be able to do this. Thanks.

Originally created by @syfbme on GitHub (Jul 5, 2021). I wonder if your model be able to do this. Thanks.
claunia commented 2026-01-29 21:36:39 +00:00
Author
Owner
Copy Link

@xinntao commented on GitHub (Jul 5, 2021):

The model can be fine-tuned to restore face w/o color. You could fine-tune it with the training codes (turn off the color augmentations in the config file).
We will upload such a model later~

@xinntao commented on GitHub (Jul 5, 2021): The model can be fine-tuned to restore face w/o color. You could fine-tune it with the training codes (turn off the color augmentations in the config file). We will upload such a model later~
claunia commented 2026-01-29 21:36:41 +00:00
Author
Owner
Copy Link

@syfbme commented on GitHub (Jul 5, 2021):

The model can be fine-tuned to restore face w/o color. You could fine-tune it with the training codes (turn off the color augmentations in the config file).
We will upload such a model later~

Thanks for your quick reply. But how to turn off the color augmentations? I only find 4 settings in file train_gfpgan_v1.yml:
color_jitter_prob: 0.3
color_jitter_shift: 20
color_jitter_pt_prob: 0.3
gray_prob: 0.01

@syfbme commented on GitHub (Jul 5, 2021): > The model can be fine-tuned to restore face w/o color. You could fine-tune it with the training codes (turn off the color augmentations in the config file). > We will upload such a model later~ Thanks for your quick reply. But how to turn off the color augmentations? I only find 4 settings in file train_gfpgan_v1.yml: color_jitter_prob: 0.3 color_jitter_shift: 20 color_jitter_pt_prob: 0.3 gray_prob: 0.01
claunia commented 2026-01-29 21:36:42 +00:00
Author
Owner
Copy Link

@syfbme commented on GitHub (Jul 5, 2021):

Do you mean that we set gray_prob to 0 so that model won't colorize it?

@syfbme commented on GitHub (Jul 5, 2021): Do you mean that we set gray_prob to 0 so that model won't colorize it?
claunia commented 2026-01-29 21:36:44 +00:00
Author
Owner
Copy Link

@xinntao commented on GitHub (Jul 5, 2021):

@syfbme
You can try:

have a very slight color jitter

color_jitter_prob: 0.5
color_jitter_shift: 5  # use a slight color jiter
color_jitter_pt_prob: ~  
gray_prob: 0.01
gt_gray: True

or
totally no color jitter

color_jitter_prob: ~
color_jitter_shift: 20
color_jitter_pt_prob: ~
gray_prob: 0.01
gt_gray: True
@xinntao commented on GitHub (Jul 5, 2021): @syfbme You can try: have a very slight color jitter ``` color_jitter_prob: 0.5 color_jitter_shift: 5 # use a slight color jiter color_jitter_pt_prob: ~ gray_prob: 0.01 gt_gray: True ``` or totally no color jitter ``` color_jitter_prob: ~ color_jitter_shift: 20 color_jitter_pt_prob: ~ gray_prob: 0.01 gt_gray: True ```
claunia commented 2026-01-29 21:36:45 +00:00
Author
Owner
Copy Link

@syfbme commented on GitHub (Jul 6, 2021):

Hi @xinntao
Thanks for your advise. But i am confused. What i want is to keep black white photo the same color without colorizing them. So i think set "gray_prob" to 0 so that model won't learn the mapping from gray to color. The settings you advised is no color jitter but the model is still able to learn the mapping from gray to color(RGB). Please correct me if i am wrong. Looking forward to your reply~

@syfbme commented on GitHub (Jul 6, 2021): Hi @xinntao Thanks for your advise. But i am confused. What i want is to keep black white photo the same color without colorizing them. So i think set "gray_prob" to 0 so that model won't learn the mapping from gray to color. The settings you advised is no color jitter but the model is still able to learn the mapping from gray to color(RGB). Please correct me if i am wrong. Looking forward to your reply~
claunia commented 2026-01-29 21:36:47 +00:00
Author
Owner
Copy Link

@xinntao commented on GitHub (Jul 6, 2021):

we also set gt_gray, which means we also make the targets to be gray. Such a setting is for better generalization for gray photos.

@xinntao commented on GitHub (Jul 6, 2021): we also set `gt_gray`, which means we also make the targets to be gray. Such a setting is for better generalization for gray photos.
claunia commented 2026-01-29 21:36:49 +00:00
Author
Owner
Copy Link

@syfbme commented on GitHub (Jul 12, 2021):

The model can be fine-tuned to restore face w/o color. You could fine-tune it with the training codes (turn off the color augmentations in the config file).
We will upload such a model later~

Hi @xinntao
When do you plan to upload the model? Looking forward to it.

@syfbme commented on GitHub (Jul 12, 2021): > The model can be fine-tuned to restore face w/o color. You could fine-tune it with the training codes (turn off the color augmentations in the config file). > We will upload such a model later~ Hi @xinntao When do you plan to upload the model? Looking forward to it.
claunia commented 2026-01-29 21:36:50 +00:00
Author
Owner
Copy Link

@syfbme commented on GitHub (Jul 19, 2021):

The model can be fine-tuned to restore face w/o color. You could fine-tune it with the training codes (turn off the color augmentations in the config file).
We will upload such a model later~

Hi @xinntao How to set the config file when fin-tuning the pretrained model.
image
There are 4 pretrained models but there are much more pretrained path in yml file:
image

How do i set each value?

Maybe just load pretrained network g and training others from beginning?

@syfbme commented on GitHub (Jul 19, 2021): > The model can be fine-tuned to restore face w/o color. You could fine-tune it with the training codes (turn off the color augmentations in the config file). > We will upload such a model later~ Hi @xinntao How to set the config file when fin-tuning the pretrained model. ![image](https://user-images.githubusercontent.com/13032160/126109663-b7df2069-fa33-42e1-8e0b-c5fcc3c6e384.png) There are 4 pretrained models but there are much more pretrained path in yml file: ![image](https://user-images.githubusercontent.com/13032160/126109851-a63f0c89-b810-4dcc-b89e-698684a1e32a.png) How do i set each value? Maybe just load pretrained network g and training others from beginning?
claunia commented 2026-01-29 21:36:51 +00:00
Author
Owner
Copy Link

@xinntao commented on GitHub (Jul 21, 2021):

@syfbme

modify the pretrain_network_g to the path to GFPGANv1.pth
And you can train others from scratch ~

@xinntao commented on GitHub (Jul 21, 2021): @syfbme modify the `pretrain_network_g` to the path to GFPGANv1.pth And you can train others from scratch ~
claunia commented 2026-01-29 21:36:52 +00:00
Author
Owner
Copy Link

@syfbme commented on GitHub (Jul 21, 2021):

@syfbme

modify the pretrain_network_g to the path to GFPGANv1.pth
And you can train others from scratch ~

I am already doing this. Since the training is slow, i will paste the result later

@syfbme commented on GitHub (Jul 21, 2021): > @syfbme > > modify the `pretrain_network_g` to the path to GFPGANv1.pth > And you can train others from scratch ~ I am already doing this. Since the training is slow, i will paste the result later
claunia commented 2026-01-29 21:36:54 +00:00
Author
Owner
Copy Link

@syfbme commented on GitHub (Jul 27, 2021):

Hi @xinntao
Current result is promising. I have set parameters as you suggested.
color_jitter_prob: ~ color_jitter_shift: 20 color_jitter_pt_prob: ~ gray_prob: 0.01 gt_gray: True
After 572k training steps, the result is as below:
image

The color is fading. However, it still colorize the below part of face. I will set the gray_prob larger(such as 0.2) to continue fine-tuning current model. Do you have any suggestions?

@syfbme commented on GitHub (Jul 27, 2021): Hi @xinntao Current result is promising. I have set parameters as you suggested. `color_jitter_prob: ~ color_jitter_shift: 20 color_jitter_pt_prob: ~ gray_prob: 0.01 gt_gray: True` After 572k training steps, the result is as below: ![image](https://user-images.githubusercontent.com/13032160/127084604-30bbfd74-ab42-46f0-9d14-fb583aa1e4a3.png) The color is fading. However, it still colorize the below part of face. I will set the gray_prob larger(such as 0.2) to continue fine-tuning current model. Do you have any suggestions?
claunia commented 2026-01-29 21:36:54 +00:00
Author
Owner
Copy Link

@xinntao commented on GitHub (Jul 27, 2021):

If you do not want to have any color enhancement, you may set color_jitter_shift: ~

@xinntao commented on GitHub (Jul 27, 2021): If you do not want to have any color enhancement, you may set `color_jitter_shift: ~`
claunia commented 2026-01-29 21:36:56 +00:00
Author
Owner
Copy Link

@syfbme commented on GitHub (Jul 27, 2021):

If you do not want to have any color enhancement, you may set color_jitter_shift: ~

I have already set "color_jitter_prob: ~", so the "color_jitter_shift" will never be used, right?

@syfbme commented on GitHub (Jul 27, 2021): > If you do not want to have any color enhancement, you may set `color_jitter_shift: ~` I have already set "color_jitter_prob: ~", so the "color_jitter_shift" will never be used, right?
claunia commented 2026-01-29 21:36:57 +00:00
Author
Owner
Copy Link

@xinntao commented on GitHub (Jul 27, 2021):

OK， you are right.

@xinntao commented on GitHub (Jul 27, 2021): OK， you are right.
claunia commented 2026-01-29 21:36:58 +00:00
Author
Owner
Copy Link

@tachikoma777 commented on GitHub (Jul 31, 2021):

The model can be fine-tuned to restore face w/o color. You could fine-tune it with the training codes (turn off the color augmentations in the config file).
We will upload such a model later~

May I ask when you plan to release pretrain mode w/o color change? Looking forward to it!

@tachikoma777 commented on GitHub (Jul 31, 2021): > The model can be fine-tuned to restore face w/o color. You could fine-tune it with the training codes (turn off the color augmentations in the config file). > We will upload such a model later~ May I ask when you plan to release pretrain mode w/o color change? Looking forward to it!
claunia commented 2026-01-29 21:37:00 +00:00
Author
Owner
Copy Link

@xinntao commented on GitHub (Aug 3, 2021):

@tachikoma777
If the training is OK, I will release it this week;-)

@xinntao commented on GitHub (Aug 3, 2021): @tachikoma777 If the training is OK, I will release it this week;-)
claunia commented 2026-01-29 21:37:02 +00:00
Author
Owner
Copy Link

@xinntao commented on GitHub (Aug 6, 2021):

@tachikoma777 @syfbme
I have updated the model without colorization. Colab demo: https://colab.research.google.com/drive/1sVsoBd9AjckIXThgtZhGrHRfFI6UUYOo

@xinntao commented on GitHub (Aug 6, 2021): @tachikoma777 @syfbme I have updated the model without colorization. Colab demo: https://colab.research.google.com/drive/1sVsoBd9AjckIXThgtZhGrHRfFI6UUYOo
claunia commented 2026-01-29 21:37:04 +00:00
Author
Owner
Copy Link

@TracelessLe commented on GitHub (Nov 11, 2021):

If you do not want to have any color enhancement, you may set color_jitter_shift: ~

Hi @xinntao @syfbme, I have set "color_jitter_prob: ~" and retrain, but the restore results of gray images are still with colorization. Would you please give some advice.

image

The config of dataset is listed below:

# dataset and data loader settings
datasets:
  train:
    name: FFHQ
    type: FFHQDegradationDataset
    dataroot_gt: ../../dataset/ffhq
    io_backend:
      # type: lmdb
      type: disk

    use_hflip: true
    mean: [0.5, 0.5, 0.5]
    std: [0.5, 0.5, 0.5]
    out_size: 256

    blur_kernel_size: 41
    kernel_list: ['iso', 'aniso']
    kernel_prob: [0.5, 0.5]
    blur_sigma: [0.1, 10]
    downsample_range: [0.8, 8]
    noise_range: [0, 20]
    jpeg_range: [60, 100]

    # color jitter and gray
    #color_jitter_prob: 0.3
    #color_jitter_shift: 20
    #color_jitter_pt_prob: 0.3
    #gray_prob: 0.01

    # If you do not want colorization, please set
    color_jitter_prob: ~
    color_jitter_shift: 0
    color_jitter_pt_prob: ~
    gray_prob: 0.05
    gt_gray: True

    crop_components: true
    component_path: experiments/pretrained_models/FFHQ_eye_mouth_landmarks_256.pth
    # eye_enlarge_ratio: 1.4

    # data loader
    use_shuffle: true
    num_worker_per_gpu: 12
    batch_size_per_gpu: 12
    dataset_enlarge_ratio: 1
    prefetch_mode: ~
@TracelessLe commented on GitHub (Nov 11, 2021): > If you do not want to have any color enhancement, you may set `color_jitter_shift: ~` Hi @xinntao @syfbme, I have set "color_jitter_prob: ~" and retrain, but the restore results of gray images are still with colorization. Would you please give some advice. ![image](https://user-images.githubusercontent.com/22371026/141266374-144c293e-f533-4aac-bcd6-cf0744bbc7bd.png) The config of dataset is listed below: ``` # dataset and data loader settings datasets: train: name: FFHQ type: FFHQDegradationDataset dataroot_gt: ../../dataset/ffhq io_backend: # type: lmdb type: disk use_hflip: true mean: [0.5, 0.5, 0.5] std: [0.5, 0.5, 0.5] out_size: 256 blur_kernel_size: 41 kernel_list: ['iso', 'aniso'] kernel_prob: [0.5, 0.5] blur_sigma: [0.1, 10] downsample_range: [0.8, 8] noise_range: [0, 20] jpeg_range: [60, 100] # color jitter and gray #color_jitter_prob: 0.3 #color_jitter_shift: 20 #color_jitter_pt_prob: 0.3 #gray_prob: 0.01 # If you do not want colorization, please set color_jitter_prob: ~ color_jitter_shift: 0 color_jitter_pt_prob: ~ gray_prob: 0.05 gt_gray: True crop_components: true component_path: experiments/pretrained_models/FFHQ_eye_mouth_landmarks_256.pth # eye_enlarge_ratio: 1.4 # data loader use_shuffle: true num_worker_per_gpu: 12 batch_size_per_gpu: 12 dataset_enlarge_ratio: 1 prefetch_mode: ~ ```
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
pull-request
Mirrored from GitHub Pull Request
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
claunia
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: TencentARC/GFPGAN#14
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?