This package simplifies DirectMHP for deployment in my project.
TODO
Using the model for inference
from DirectMHP_Inference_seoy import DirectMHP
model = DirectMHP(weights='path/to/weights', device=device)To save model in onnx format
from DirectMHP_Inference_seoy import save_with_weights
save_with_weights(weights='path/to/weights', device='device')or run the save_model script in the package
- YOLOv5 🚀 in PyTorch > ONNX > CoreML > TFLite
- BMVC 2020 - WHENet: Real-time Fine-Grained Estimation for Wide Range Head Pose
- CVPR 2021 - img2pose: Face Alignment and Detection via 6DoF, Face Pose Estimation
- ICIP 2022 - 6D Rotation Representation for Unconstrained Head Pose Estimation
- We also thank public datasets AGORA and CMU-Panoptic for their excellent works.
Our work is based on public code and datasets. If you plan to add our work to your business project, please obtain the following enterprise licenses.
- DirectMHP: GNU General Public License v3.0 (GPL-3.0 License): See LICENSE file for details.
- YOLOv5: To request an Enterprise License please complete the form at Ultralytics Licensing
- AGORA-HPE: Data & Software Copyright License for non-commercial scientific research purposes AGORA License
- CMU-HPE: CMU Panoptic Studio dataset is shared only for research purposes, and this cannot be used for any commercial purposes. The dataset or its modified version cannot be redistributed without permission from dataset organizers CMU Panoptic Homepage