Blender Data Generation

This is the official repository for the paper : C3I-SynFace: A synthetic head pose and facial depth dataset using seed virtual human models.

• Import FBX from iClone
• Import Missing Files in Blender
• Scene Setup in Blender
• Compositor Setup in Blender
• Render Ground Truth with Cycle Render Engine

Additionally -

• Data Preparation Script

• Neck Bone Adjustment through Iterative Script

• Blender Scenes can be downloaded from - https://drive.google.com/file/d/18UF2DQpHQ5N7L9Z-LCcD4P_elj_-LDco/view?usp=sharing

Install package within Blender Python

• Open Blender python editor

• Locate the blender python version

  Normally in the <Blender_Root>/<Blender_Version>2.93/python/bin

  Run the following scripts from Blender python terminal -

# directory - face3d/mesh/cython
import sys 
import os 
import subprocess 
  

python_exe = os.path.join(sys.prefix, 'bin','python3.9')

# Upgrade Pip
subprocess.call([python_exe, "-m", "ensurepip"]) 
subprocess.call([python_exe, "-m", "pip", "install", "--upgrade", "pip"]) 

# Install sympy
subprocess.call([python_exe, "-m", "pip", "install", "sympy"]) 

Dataset

The raw dataset can be found in the following path (alternate to Mendeley) -

3D full body model -

https://drive.google.com/drive/folders/177Xem5rLg7GYRn6IDwWMwZtBgr57OrtB?usp=share_link

Head Pose dataset -

https://drive.google.com/drive/folders/10QNIb4Rp9D7SHMbdiK3ecbZFIL_bNOEY?usp=share_link

Face depth Dataset -

https://drive.google.com/drive/folders/1oleqLbR793xBmw8gF91JTi4TrBJQUMr2?usp=share_link

Citation

If you find our work useful to your research, please consider citing:

@article{basak2023c3i,
  title={C3I-SynFace: A synthetic head pose and facial depth dataset using seed virtual human models.},
  author={Basak, Shubhajit and Khan, Faisal and Javidnia, Hossein and Corcoran, Peter and McDonnell, Rachel and Schukat, Michael},
  journal={Data in Brief},
  pages={109087},
  year={2023},
  publisher={Elsevier}
}