A Tool to unfold RenderMe360 data

A simple tool to unfold .smc files in the RenderMe360 data into a readable and usable format.

⚙️  Install

• python base
  • Python 3.9 or higher
• Other Packages:

pip install opencv-python h5py numpy tqdm pydub plyfile

To download the RenderMe360 data, you need to install openxlab library using pip install -U openxlab

 Usage

You can either test some frame data in some view using test_data.py or unfold all data of one actor using unfold_data.py.

Don't forget change '/path/to/RenderMe360' to your own dataset path.

Some modifiable parameters

• DATA_ROOT: the root path of your RenderMe360 data.
• ACTOR_ID: the actor index which you want to test or unfold.
• UNFOLD_LIST: identify which items you want to unfold. (You can choose from these items: "image", "mask", "uv", "scan", "lmk_2d", "lmk_3d", "audio")
• SKIP_SEQ: skip some expressions, speeches, or hairstyles.