3D Strand-Based Hair Generation and Editing System: Using HairStep to Generate Hair from Images and Export Editable and Applicable Files

Requirements

Anaconda 3, Blender 4.3

Installation & Execution

git clone --recursive https://github.com/asd56585452/3D-hair-reconstruction-interface.git

cd 3D-hair-reconstruction-interface

cd HairStep

conda env create -f environment.yml
conda activate hairstep

pip install torch==1.9.0+cpu torchvision==0.10.0+cpu torchaudio==0.9.0 -f https://download.pytorch.org/whl/torch_stable.html

pip install -r requirements.txt

cd external/3DDFA_V2
sh ./build.sh
cd ../../
cd ..

rm -r HairStep/results
pip install PyQt6==6.2.3
pip install pyrr
pip install PyOpenGL
cp -r HairStepCpuPatch/* HairStep
cd grid_less

sudo apt-get install libxcb-cursor0

Next, download the model files from the HairStep GitHub repository.

Download the SAM checkpoint and place it in ./HairStep/checkpoints/SAM-models/.

Download the 3D networks checkpoint and place it in ./HairStep/checkpoints/recon3D/.

Then, execute the following command (make sure to run the command in the ./grid_less/ folder):

python grid_slicer.py

The code has been tested on Ubuntu 24.04 LTS.

If you encounter an "ld" error while running "sh ./build.sh", it could be due to an issue with anaconda3/envs/hairstep/compiler_compat/ld. You can delete it. Refer to:

• ninia/jep#446
• ContinuumIO/anaconda-issues#11152

Demo and Usage Examples

Google Drive link