https://github.com/calciferzh/minimal-body
A very simple baseline to estimate 2D & 3D SMPL-compatible keypoints from a single color image.
https://github.com/calciferzh/minimal-body
3d-pose-estimation computer-vision deep-learning
Last synced: about 1 year ago
JSON representation
A very simple baseline to estimate 2D & 3D SMPL-compatible keypoints from a single color image.
- Host: GitHub
- URL: https://github.com/calciferzh/minimal-body
- Owner: CalciferZh
- License: mit
- Created: 2021-11-19T14:16:40.000Z (over 4 years ago)
- Default Branch: master
- Last Pushed: 2021-11-20T11:18:04.000Z (over 4 years ago)
- Last Synced: 2025-04-01T15:11:53.718Z (over 1 year ago)
- Topics: 3d-pose-estimation, computer-vision, deep-learning
- Language: Python
- Homepage:
- Size: 290 KB
- Stars: 56
- Watchers: 4
- Forks: 5
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Minimal Body
A very simple baseline to estimate 2D & 3D SMPL-compatible keypoints from a single color image.
The model file is only 51.2 MB and runs at 70fps on a 2080Ti.
## Usage
1. Download the pre-trained model [here](https://github.com/CalciferZh/minimal-body/releases/download/v1/minimal_body_v1.pth) into `model/`.
2. Install the dependencies.
3. Run `python example.py`.
The input image should already be cropped to a 4:3 portrait with the subject in the center.
## Misc
We will __not__ release the training code.
The model is trained using the following datasets jointly:
* [COCO](https://cocodataset.org/)
* [MPII2D](http://human-pose.mpi-inf.mpg.de/)
* [MPII3D](https://vcai.mpi-inf.mpg.de/3dhp-dataset/)
* [HUMBI](https://humbi-data.net/)
* [HM36M](http://vision.imar.ro/human3.6m/description.php)
* [MTC](http://domedb.perception.cs.cmu.edu/mtc.html)
* [SPIN](https://github.com/nkolot/SPIN)
Please also check the license of the listed datasets.
If you find it helpful, please consider citing our related paper:
```
@InProceedings{Zhou_2021_CVPR,
author = {Zhou, Yuxiao and Habermann, Marc and Habibie, Ikhsanul and Tewari, Ayush and Theobalt, Christian and Xu, Feng},
title = {Monocular Real-Time Full Body Capture With Inter-Part Correlations},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2021},
pages = {4811-4822}
}
```