Swapping only one dancer in a moving multi-person video with Wan 2.2 Animate
asduskun · reddit · 2026-10-04
Using Wan 2.2 Animate in ComfyUI on RunPod, the poster wants to replace only the main dancer in a handheld nighttime video while keeping everyone else intact. Workflow: WanVideoWrapper, GGUF Q3 model, SAM2 mask video, DWPose, reference image. Without masking, characters and poses blend; with SAM2, the target sometimes gets covered by a black mask or removed instead of replaced. Blockify Mask, Mask Grow, and fixed crops all fail due to camera and subject movement. They ask about mask wiring/inversion, correct DWPose input, and better single-person tracking nodes.
More from Multimodal
- Grok Imagine quietly adds intermediate keyframes for video generation — mark_k · 2026-10-04
- Google quietly tests new Gemini 2.5 Flash image generation model on Arena — mornaji · 2026-10-04
- W.I.T.C.H. fan trailer made 95% locally: full AI video pipeline on a 4070 + 5060ti — No_Taste_4102 · 2026-10-04
- Balancing inpainting cadence in video: more frequent inpaints mean less smear but more hectic detail — matdryhurst · 2026-10-04
- User claims Suno has achieved 'musical AGI', composing better than 95% of people — astralmatrix · 2026-10-04
- "Bad Footage" Makes AI Video More Believable: A UFO Vlog Demo with Full Prompt — LudovicCreator · 2026-10-04