DECA

DECA copied to clipboard

Thanks a lot for your great work. I wonder how can I feed the model with images of a human face captured from multiple views (at the same time of...

ShaYito

Extracting shape, pose, expression parameters from meshes (3d or 2.5d, not 2d image)

4

Hello, thanks for your awesome work. Is it possible to extract the shape, pose, expression parameters from meshes, not from 2d images? In other words, can the meshes be input?...

bbakpil

which software can be used to open this mesh other than meshlab

1

# as i'm openning this mesh in another software im seeing like this  # how to solve this problem ??

zardamhussain

Steps for setting up the environment in new GPU like 4090

4

PS: This is not an issue. Adding it here as the fellow community dont need spend too much time in setting up the env **Anaconda || Ubuntu 22.04 || Python...

epratheeban

No module named 'pytorch3d.structures'

1

/usr/local/lib/python3.10/dist-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /usr/local/lib/python3.10/dist-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or...

Garv62003

/usr/local/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead. warnings.warn( /usr/local/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or...

Garv62003

I set `--iscrop=false`, But the person's face is looking down. What's going on? 

Echo-jyt

``` ## projection landmarks2d = util.batch_orth_proj(landmarks2d, codedict['cam'])[:,:,:2]; landmarks2d[:,:,1:] = -landmarks2d[:,:,1:]#; landmarks2d = landmarks2d*self.image_size/2 + self.image_size/2 landmarks3d = util.batch_orth_proj(landmarks3d, codedict['cam']); landmarks3d[:,:,1:] = -landmarks3d[:,:,1:] #; landmarks3d = landmarks3d*self.image_size/2 + self.image_size/2 trans_verts =...

Echo-jyt

Got the colorless texture and false training Detail for mesh

2

  The first image show the false mesh after taking the detail training (I tested that it came from photo_detail_mrf loss.) The second one, it show the colorless...

zouiner

I'm coming here from CodeTalker repo since it needs .pkl and .ply files both as output.

pillai-karthik