-
Notifications
You must be signed in to change notification settings - Fork 8
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
Performance about normal map #11
Comments
Hi, what is the normal maps in the training set like? |
CUDA_VISIBLE_DEVICES= python train_3d.py --data="/home/youzehong/jt/blender-3.2.2-linux-x64/GETAvatar/datasets/THuman2.0/THuman2.0_res512/" --gpus=1 --batch=32 --batch-gpu=1 --mbstd-group=1 --gamma=10 --dmtet_scale=2 --one_3d_generator=1 --fp32=0 --img_res=512 --norm_interval=1 --dis_pose_cond=True --normal_dis_pose_cond=True --eik_weight=1e-3 --unit_2norm=True --use_normal_offset=False --blur_rgb_image=False --blur_normal_image=False --camera_type=blender --load_normal_map=True --with_sr=True --outdir="/home/youzehong/jt/GETAvatar-main/thuman_res512_ckpts/" |
What is the result if using ''--gpus=8 --batch=32 --batch-gpu=4 --mbstd-group=4''? |
Hahaha, I mainly don't have so many GPU resources, thank you for your opinion! |
Maybe you can increase the gamma value to see if the result is better. |
Thanks, I will try it.
…---- Replied Message ----
| From | Xuanmeng ***@***.***> |
| Date | 12/12/2023 19:59 |
| To | magic-research/GETAvatar ***@***.***> |
| Cc | somnus1840 ***@***.***>,
Author ***@***.***> |
| Subject | Re: [magic-research/GETAvatar] Performance about normal map (Issue #11) |
Maybe you can increase the gamma value to see if the result is better.
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you authored the thread.Message ID: ***@***.***>
|
According to the provided training instructions, the training effect of the normal map is much worse than that of the texture. May I ask what is the specific reason or are there any ways to improve it? Thanks very much!
!
The text was updated successfully, but these errors were encountered: