ID-Sculpt

Preprocessing Pipeline

Environment

Environment setup refers to configurations in Panohead and 3DDFA_V2

One-click Processing Script

bash preprocess.sh

Processing Pipeline

1. Face alignment (working directory is under 3DDFA_V2-master folder, images in input folder should be .png)

   cd 3DDFA_V2-master
   python dlib_kps.py -input_dir $DIR -output_path $PATH
   # DIR = test/head_img_new, PATH = data.pkl
   

2. Calculate 3DMM and camera parameters (dataset.json: image names, camera parameters; face_data.pkl: coordinates of seven facial keypoints)

   cd 3DDFA_V2-master
   python recrop_images.py -i data.pkl -j dataset.json --out_dir '../PanoHead-main/dataset/head_img_new'
   

3. get rgba image

   cd Panohead/face_parsing
   python evaluate.py  -input_dir $DIR -output_path $DIR
   

4. run diffusion inference to get depth and normal

   cd preprocess
   python get_facedepth.py
   python get_facenormal.py
   

dataset：/home/haojinkun/3DAvatar/PanoHead-main/dataset/head_img_new

1. run panohead inversion, and get triplane weight

   cd Panohead
   bash gen_pti_script.sh
   

2. Export multi-view normal maps and generate mesh .ply file

   bash gen_mesh.sh
   

3. Align landmark

   python pre_align_3Dlm.py