i2i2v (i2i + i2v) - AIO Qwen 2511 Background Removal + Wan 2.2 360 Video/Batch Generator — i2i2v BG Remove/360 v1

Use an image of a person to create a white-background subject image, a Wan 2.2 360-orbit video, and a matching frame batch for character creation, references, or LoRA training.

The Qwen 2511 stage removes the background and places the subject on white; Wan 2.2 uses that image to create the video and frame batch.

You can queue multiple images and runs through the combined stages.

Start with an image of a person; optional second and third images can serve as references.

Outputs include a clean image on a solid white background, a 360-orbit video, and a batch of images corresponding to its frames.

Have the required ComfyUI nodes and model files named in the workflow available, choosing models suited to your VRAM, RAM, and CPU.

Suggestion · non vérifié

For a single subject, keep the built-in prompts unless you need a specific pose, angle, or use case; if a result is poor, try a different seed before heavily changing them.

Suggestion · non vérifié

Set the Wan 2.2 video output resolution and frame rate to suit your needs.

Suggestion · non vérifié

Use the optional second and third image inputs for references or the optional rgthree LoRA node for an in-transit edit; if the subject does not spin, check that an Orbit LoRA is in use.

Suggestion · non vérifié

Wan 2.2 can struggle with multiple subjects, elaborate poses, or unusual rotation axes, and may produce ghosting, artifacts, or extra limbs.

Very large images took substantially longer in the author's testing; object and item prompts remain WIP and need broader testing.