vt-vl-lab / 3d-photo-inpainting

[CVPR 2020] 3D Photography using Context-aware Layered Depth Inpainting

Home Page:https://shihmengli.github.io/3D-Photo-Inpainting/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Alternative repo?

szriru opened this issue · comments

commented

It seems this repo isn't maintained anymore, and have full of dependency errors.
Could you provide me alternative apps/softwares to create moving 3d photos but locally and high resolutions?

I tried this repo and other forked repos that updated recently in windows & wsl ubuntu, but couldn't make it run without errors.

commented

Somehow I manged to make it work. So I will write down what i did in short.
I used wsl Ubuntu in windows 11.

follow the instruction of this repo to the end and

  • Make sure you installed CUDA in windows
  • git clone this repo
  • in main.py. change the line 29. yaml.load ===> yaml.full_load

To make ubuntu recognize CUDA in windows
Run CMD in Windows (as Administrator)

C:
cd \Windows\System32\lxss\lib
del libcuda.so
del libcuda.so.1
mklink libcuda.so libcuda.so.1.1
mklink libcuda.so.1 libcuda.so.1.1

torch has bugs to ignore GPU, so downgrade it.

pip install torch==1.10.1+cu111 torchvision==0.11.2+cu111 -f https://download.pytorch.org/whl/torch_stable.html

install xvfb

sudo apt-get install xvfb
pip install xvfbwrapper

every time you launch ubuntu you should do this

export MKL_SERVICE_FORCE_INTEL=1
export MKL_THREADING_LAYER=GNU

and

sudo apt-get install -y mesa-utils and libgl1-mesa-glx
conda install -c conda-forge libstdcxx-ng
pip install moviepy --upgrade

This is just like a memo after all.
sry for that but i hope this will help someone.

bro how can i use it in google colab i dont have a beast pc?

I don't know.
You could use this : https://convert.leiapix.com/animation

can you try to use this repo on google colab once please i realy need your code to work dude i cant use leiapix

I don't use google colab.
You should give it a try by your own. Good luck.