afruehstueck / insetgan Goto Github PK

RuntimeError: CUDA out of memory. Tried to allocate 1.50 GiB (GPU 0; 8.00 GiB total capacity; 4.47 GiB alrea

Hi! May I ask you two questions about G_canvas and D_canvas?

1. In the code I found G_canvas is defined but D_canvas is not defined. So how to solve this problem?
2. In the paper, I didn't read anything about discriminator in the paper. But why something about discriminator appears in the code?
   Thank you.

Multimodal Body Generation for an Existing Face.
For each face generated by the pretrained FFHQ model (middle
column), we use joint optimization to generate three different bodies while maintaining the facial identities from the input faces

code:
with open('./networks/DeepFashion_1024x768.pkl', 'rb') as f:
data = pickle.load(f)
dir_path = data['training_set_kwargs']['path']

Where is this path - '/tmp/fruehsa/data/FashionHumans.zip' ? Which folder should I go to ?
Thank you!

Hey,
Thanks for such greate work!
I want to ask for the shoes insetGAN pre-trained model
Or at least what are the instructions for shoes inset adaption

ResolvePackageNotFound when downloading, versions cannot be found.

In my view, the InsetGAN was built on the Generator trained from the StyleGAN-v2.

In other words, the InsetGAN only need to use the trained StyleGAN-v2 to generate thousands of images of bodys and faces, so that, the InsetGAN could insert the inset into the whole-body image, and make it a coherent whole-body image with face adjusted.

So the InsetGAN is just a image-mix network which has no need to "train it" .It is just a effective tool.

Then the part of “Generate your own Human Dataset” is used for show the mix result on my own human data which would be used to train on the StyleGAN-v2.

Am I right? Appreciate it if u could answer.

Can I use insetGAN for headtransfer? That is, take the head and neck part of user1 and seamlessly fit it on the body of user 2?

In my view, the InsetGAN was built on the Generator trained from the StyleGAN-v2.

In other words, the InsetGAN only need to use the trained StyleGAN-v2 to generate thousands of images of bodys and faces, so that, the InsetGAN could insert the inset into the whole-body image, and make it a coherent whole-body image with face adjusted.

So the InsetGAN is just a image-mix network which has no need to "train it" .It is just a effective tool.

Then the part of “Generate your own Human Dataset” is used for show the mix result on my own human data which would be used to train on the StyleGAN-v2.

Am I right? Appreciate it if u could answer.

Hello, may I ask if you could share the engineering file of using FID to evaluate images

Can you illustrate for the provided pre-trained canvas generator model， which subset of deepfashion dataset is used for training?

Under which py file can set batchsize?
Thank you!

Hi. I have read your paper in CVPR, and I have some question about this project.

In paper, the two regularization terms are
$L_{reg}=\left| w^{*} - w_{avg}\ \right|+\sum \left| \delta_i \right|$

I only can find the first term. Since the second term is not used in the project, I don't know how it works about δ.

absl-py 0.15.0
astunparse 1.6.3
cachetools 5.3.0
certifi 2022.12.7
charset-normalizer 3.1.0
cloudpickle 2.2.1
colorama 0.4.6
cycler 0.11.0
dask 2022.2.0
dnnlib 0.0.1
facenet-pytorch 2.5.2
ffmpeg-python 0.1.17
flatbuffers 1.12
fonttools 4.38.0
fsspec 2023.1.0
future 0.18.3
gast 0.3.3
google-auth 2.17.3
google-auth-oauthlib 0.4.6
google-pasta 0.2.0
grpcio 1.32.0
h5py 2.10.0
idna 3.4
importlib-metadata 6.6.0
Keras-Preprocessing 1.1.2
kiwisolver 1.4.4
locket 1.0.0
lpips 0.1.4
Markdown 3.4.3
MarkupSafe 2.1.2
matplotlib 3.5.3
networkx 2.6.3
numexpr 2.8.4
numpy 1.21.6
oauthlib 3.2.2
onnx 1.14.0
opt-einsum 3.3.0
packaging 23.1
partd 1.4.0
Pillow 9.5.0
pip 22.3.1
protobuf 3.20.3
pyasn1 0.5.0
pyasn1-modules 0.3.0
pyparsing 3.0.9
PyQt5 5.15.9
PyQt5-Qt5 5.15.2
PyQt5-sip 12.12.1
python-dateutil 2.8.2
PyWavelets 1.4.0
PyYAML 6.0
requests 2.30.0
requests-oauthlib 1.3.1
rsa 4.9
scikit-image 0.14.2
scipy 1.4.1
setuptools 65.6.3
six 1.15.0
tensorboard 2.11.2
tensorboard-data-server 0.6.1
tensorboard-plugin-wit 1.8.1
termcolor 1.1.0
tf2onnx 1.9.3
toolz 0.12.0
torch 1.10.0
torch-utils 0.1.2
torchaudio 0.10.0
torchvision 0.11.1
tqdm 4.65.0
typing-extensions 3.7.4.3
urllib3 2.0.2
Werkzeug 2.2.3
wheel 0.38.4
wrapt 1.12.1
zipp 3.15.0

python run_insetgan.py

Traceback (most recent call last):
File "run_insetgan.py", line 40, in
networks = pickle.Unpickler(f).load()
ModuleNotFoundError: No module named 'torch_utils.persistence'

Can the version of torch and cuda be higher than the required version? And although higher versions need to be specified? Could you give some examples of torch and cuda versions?
Thank you!