| # easy_ViTPose | |
| <p align="center"> | |
| <img src="https://user-images.githubusercontent.com/24314647/236082274-b25a70c8-9267-4375-97b0-eddf60a7dfc6.png" width=375> easy_ViTPose | |
| </p> | |
| ## Accurate 2d human pose estimation, finetuned on 25 keypoints COCO skeleton + feet | |
| ### Easy to use SOTA `ViTPose` [Y. Xu et al., 2022] models for fast inference. | |
| These are just the models, refer to https://github.com/JunkyByte/easy_ViTPose for the actual code. | |
| | Models | Path | | |
| | :----: | :----: | | |
| | TORCH | [Folder](https://huggingface.co/JunkyByte/easy_ViTPose/tree/main/torch) | | |
| | ONNX | [Folder](https://huggingface.co/JunkyByte/easy_ViTPose/tree/main/onnx) | | |
| | TENSORRT | [Folder](https://huggingface.co/JunkyByte/easy_ViTPose/tree/main/tensorrt) | | |
| You can also download the YOLOv5 models: | |
| | Models | Path | | |
| | :----: | :----: | | |
| | YOLOv5 | [Folder](https://huggingface.co/JunkyByte/easy_ViTPose/tree/main/yolov5) | | |
| ### License | |
| Refer to official https://github.com/ViTAE-Transformer/ViTPose/blob/main/LICENSE for model license |