Brooooo What to do 🙁 Am getting the same error even tested with different pictures with simple backgrounds 🙁 Even after editing in photoshop with simple T poses. net = PoseEstimationWithMobileNet() checkpoint = torch.load('checkpoint_iter_370000.pth', map_location='cpu') load_state(net, checkpoint) get_rect(net.cuda(), [image_path], 512)
The program code gives an error. ERROR: Could not find a version that satisfies the requirement torch ---- ERROR: No matching distribution found for torch==1.6.0+cu101
As it is already trained on tons of data about people wearing different kinds of clothes, it understands how to "estimate" the backside, we can see that the clarity of the backside is low as a result!
Please make a simple guide how to download this, make your own 3d photos or 3d videos using home pc and RTX GPU. Without the need for Google Collab or anything online :) Would be great to generated 3D Side by Side images from 2D images.
Run PIFuHD! Resuming from ./checkpoints/pifuhd.pt Warning: opt is overwritten. test data size: 1 initialize network with normal initialize network with normal generate mesh (test) ... 0% 0/1 [00:00
Brooooo What to do 🙁 Am getting the same error even tested with different pictures with simple backgrounds 🙁 Even after editing in photoshop with simple T poses.
net = PoseEstimationWithMobileNet()
checkpoint = torch.load('checkpoint_iter_370000.pth', map_location='cpu')
load_state(net, checkpoint)
get_rect(net.cuda(), [image_path], 512)
Amazing work.
any exe file ? github files do not work for me -.- its crazy complicated to start it
The program code gives an error. ERROR: Could not find a version that satisfies the requirement torch ---- ERROR: No matching distribution found for torch==1.6.0+cu101
PIFu model is applicable only for the image of human and not for objects (eg., image of chair)?
Is this only for human bodies?
How does it calculate the clothes or whatever it is in the back of a 2d data!?
As it is already trained on tons of data about people wearing different kinds of clothes, it understands how to "estimate" the backside, we can see that the clarity of the backside is low as a result!
Please make a simple guide how to download this, make your own 3d photos or 3d videos using home pc and RTX GPU. Without the need for Google Collab or anything online :) Would be great to generated 3D Side by Side images from 2D images.
Where can we find the software to try it out ourselves?
Colab:
colab.research.google.com/drive/11z58bl3meSzo6kFqkahMa35G5jmh2Wgt
GitHub:
github.com/facebookresearch/pifuhd
Didnt work came out on blender distorted.
I'm getting an Attribute Error even though I followed your tutorials completely
interesting!
同じ名前です
Run PIFuHD!
Resuming from ./checkpoints/pifuhd.pt
Warning: opt is overwritten.
test data size: 1
initialize network with normal
initialize network with normal
generate mesh (test) ...
0% 0/1 [00:00