The project adopts the KJ wan1.3B workflow. Project address: https://github.com/kijai/ComfyUI WanVideoWrapper

It allows control over whether to enable blurring to adjust the degree of redrawing,

Enable blurring: This will significantly alter the details of the source video, suitable for scenarios like hand and face detail repair.

Disable blurring: This can refine details based on the source video, suitable for redrawing landscapes, distant figures, and other scenes.

Blurring and non-blurring can be used in combination for processing. For example: First, enable blurring to complete face and hand repairs, then use the output video as input to disable blurring for further detail refinement.

It is recommended to use AI-assisted generation for prompts to achieve better results. The description should be similar to the source video, such as: A 25-year-old girl (main subject description), tall and wearing a red gown, with the hem of the dress swaying in the sea breeze (main subject dynamics), barefoot dancing on the beach (motion description). The dance is elegant and rhythmic, with arms stretched like a swan spreading its wings, and the waist twisting to make the dress hem rotate (motion details).

Based on personal testing, this approach works better for videos with small motion amplitudes. For videos with excessive motion amplitudes, the repair effect is suboptimal...

Creating content is not easy. If this has helped you, please save, like, execute applications or workflows online. Your support is my motivation for creating. Thank you for your support!

If you have any questions, please leave a comment. I will reply promptly when I see it. Feel free to engage, learn together, and make progress together!