ComfyUI-WanAnimatePreprocess

WanAnimatePreprocess

• ONNX Detection Model Loader

  Loads ONNX models for pose and face detection. ViTPose for pose estimation and YOLO for object detection.

• Pose and Face Detection

  Detects human poses and face images from input images. Optionally retargets poses based on a reference image.

• Pose Detection OneToAll Animation

  Specialized pose detection and alignment for OneToAllAnimation model https://github.com/ssj9596/One-to-All-Animation. Detects poses from input images and aligns them based on a reference image if provided.

• Pose Retarget Prompt Helper

  Generates text prompts for pose retargeting based on visibility of arms and legs in the template pose. Originally used for Flux Kontext