JunukCha / MultiPerson

Multi-Person 3D Pose and Shape Estimation via Inverse Kinematics and Refinement

MultiPerson

Official code of Multi-Person 3D Pose and Shape Estimation via Inverse Kinematics and Refinement

Junuk Cha, Muhammad Saqlain, GeonU Kim, Mingyu Shin, Seungryul Baek

In the Wild Input	Output

Installation

git clone https://github.com/JunukCha/MultiPerson.git

cd MultiPerson

Anaconda must be installed in your computer.

source install.sh

If it has been installed well so far, it has a folder tree structure as shown below and your virtual environment name is set to 'MultiPerson'.

.MultiPerson
├── configs
├── demo_image
├── lib
├── .gitignore
├── README.md
├── demo.py
├── install.sh
├── requirements.txt
└── YOLOv4

Data Preparation

For Yolo, download yolov4.pth and place it as below.

.MultiPerson
└── YOLOv4
    └── weight
        └── yolov4.pth

1. Download data in Google Drive and place them as below.
2. Download "basicModel_neutral_lbs_10_207_0_v1.0.0.pkl" from SMPL web page and place it as below.

.MultiPerson
├── data
│   ├── base
│   │   ├── 32_to_122.npy
│   │   └── joint_info.pkl
│   ├── checkpoints
│   │   ├── inverse_kinematics.pth
│   │   ├── model_checkpoint.pt
│   │   ├── pose_estimator.pth
│   │   └── transformer.pth
│   └── smpl
│       ├── basicModel_neutral_lbs_10_207_0_v1.0.0.pkl
│       ├── h36m_mean_beta.npy
│       ├── J_regressor_extra.npy
│       ├── J_regressor_h36m.npy
│       └── smpl_mean_params.npz

In 'yolov4.pth', the key name 'neek' of the parameter should be changed to 'neck'.

python change_param_key.py

Inference

python demo.py --img demo_image/demo1.jpg

Citation

@inproceedings{cha2022multi,
  title={Multi-Person 3D Pose and Shape Estimation via Inverse Kinematics and Refinement},
  author={Cha, Junuk and Saqlain, Muhammad and Kim, GeonU and Shin, Mingyu and Baek, Seungryul},
  booktitle={ECCV},
  year={2022},
}