SDPose: Exploiting Diffusion Priors for Out-of-Domain and Robust Pose Estimation
📄 Paper
💻 GitHub
🤗 Body Model
🤗 WholeBody Model
⚙️ Settings
Model Selection
Body (17 kpts) or WholeBody (133 kpts)
body
wholebody
Model Status
Select model and upload media
Enable YOLO Detection
For multi-person detection
Confidence Threshold
↺
0.1
0.9
Pose Overlay Opacity
↺
0
1
📷 Image
🎬 Video
📷 Image
🎬 Video
Input Image
Drop Image Here
- or -
Click to Upload
Output with Keypoints
📥 Download JSON
Detection Results
🔍 Run Image Inference
📥 Download Processed Video
Processing Results
🎬 Run Video Inference
📝 Usage
Select model (Body or WholeBody)
Upload image or video
Configure settings
Click Run button
Download results
⚠️ Notes
First load may take 1-2 minutes
YOLO-det recommended for multi-person
Video processing may be slow on CPU