Python Embeded for GVHMR
Patreon
If you are a Patron of CEB Studios, please contact me so I can make you a discount code.
WARNING:
Be sure that your computer have the necessary requerements to run.
It was tested on a PC with RTX 3090 24gb of VRAM and 48GB of RAM
It should run on lower settings, but I cant say what is the minimum.
WARNING 2:
The result from GVHMR should not be used commercially due to license from GVHRM
https://github.com/zju3dv/GVHMR?tab=License-1-ov-file#readme
and SMPL/SMPLX
https://smpl.is.tue.mpg.de/modellicense.html
https://smpl-x.is.tue.mpg.de/modellicense.html
Video Tutorial showing the process
Instructions
This is only the python embeded with the python packages necessary to run GVHMR
You still need to download the GVHMR code from the github page :
https://github.com/carlosedubarreto/GVHMR (this repository already has some tools to generate the data to use in Blender, but the original repository should work too)
Download the checkpoints from: https://drive.google.com/drive/folders/1eebJ13FUEXrKBawHpJroW0sNSxLjh9xD
SMPL
Download SMPL from: https://smpl.is.tue.mpg.de/
(direct link: https://download.is.tue.mpg.de/download.php?domain=smpl&sfile=SMPL_python_v.1.1.0.zip)
Download SMPL FBX model (to use in Blender)
(Direct link: https://download.is.tue.mpg.de/download.php?domain=smpl&sfile=SMPL_maya.zip)
SMPLX
Download SMPLX from https://smpl-x.is.tue.mpg.de/ (direct link, https://download.is.tue.mpg.de/download.php?domain=smplx&sfile=models_smplx_v1_1.zip)
Place the SMPL, SMPLX and checkpoints inside the folder inputs/checkpoints , according to the instructions on the page https://github.com/zju3dv/GVHMR/blob/main/docs/INSTALL.md#inputs--outputs
The structure you should add the files
SMPL, SMPLX and the checkpoints
# 1. You need to sign up for downloading [SMPL](https://smpl.is.tue.mpg.de/) and [SMPLX](https://smpl-x.is.tue.mpg.de/). And the checkpoints should be placed in the following structure:
inputs/checkpoints/
├── body_models/smplx/
│ └── SMPLX_{GENDER}.npz # SMPLX (We predict SMPLX params + evaluation)
└── body_models/smpl/
└── SMPL_{GENDER}.pkl # SMPL (rendering and evaluation)
# 2. Download other pretrained models from Google-Drive (By downloading, you agree to the corresponding licences): https://drive.google.com/drive/folders/1eebJ13FUEXrKBawHpJroW0sNSxLjh9xD?usp=drive_link
inputs/checkpoints/
├── dpvo/
│ └── dpvo.pth
├── gvhmr/
│ └── gvhmr_siga24_release.ckpt
├── hmr2/
│ └── epoch=10-step=25000.ckpt
├── vitpose/
│ └── vitpose-h-multi-coco.pth
└── yolo/
└── yolov8x.ptPlace the GVHMR folder from the github in the same folder as the python_embeded folder, like this
Where the unziped GVHMR files are in a folder like this:
And one last thing, you'll have to run the command to install gvhmr
Following the folder structure from the example, you should be at: D:\AI\GVHMR\GVHMR to run the command
..\python_embedded\python -m pip install -e .The result should be something like this:
Now, you can run GVHMR ( you need to be in the same folder that you ran the previous command , from the example, it would be in D:\AI\GVHMR\GVHMR:
..\python_embedded\python tools/demo/demo.py --video=docs/example_video/tennis.mp4 -sThe result might be something like this
If you wanna talk more about, come to our discord
- discord server: https://discord.gg/BRuu43Nv2J