Realistic 3D Avatar Pipeline using PIFuHD - Pixel Aligned Implicit Functions and RigNet - Neural Rigging Network
This project was tested in Windows 10 x64 System with Python 3.7 | Torch v1.8.0 TorchVision v0.9.0 on CPU with 16Gb of RAM using Anaconda Python Environment.
open3D supports only Python v3.6, 3.7 and 3.8
conda create -n 3D_Avatar_Pipeline python=3.7
conda activate 3D_Avatar_Pipeline
git clone https://github.com/codesavory/3d_avatar_pipeline
git submodule init
git submodule update
pip install torch==1.8.0 torchvision==0.9.0
pip install -r requirements.txt
conda install pytorch==1.8.0 torchvision==0.9.0 torchaudio==0.8.0 cudatoolkit=10.2 -c pytorch
pip install torch-scatter -f https://pytorch-geometric.com/whl/torch-1.8.0+cu101.html
pip install torch-sparse -f https://pytorch-geometric.com/whl/torch-1.8.0+cu101.html
pip install torch-cluster -f https://pytorch-geometric.com/whl/torch-1.8.0+cu101.html
pip install torch-spline-conv -f https://pytorch-geometric.com/whl/torch-1.8.0+cu101.html
pip install torch-geometric
Download Windows-compiled Rtree from here, and install it by
pip install Rtree‑0.9.4‑cp37‑cp37m‑win_amd64.whl
(64-bit system) or
pip install Rtree‑0.9.4‑cp37‑cp37m‑win32.whl
(32-bit system). Other libraries can be installed in the same way as Linux setup instructions.
Download the checkpoints into the base folder from GDrive Link extract and rename it as Checkpoints
The given script takes input photo(from ./input folder) and stores all the results(to ./Results/). Example usage -
python 3D_Avatar_Pipeline.py .\input\test.png