First version release.
Profil d'exécution
Description de la source
❤️ If you enjoy my work and would like to support me, consider buying me a ☕ Coffee .
Apr-10-25 hotfix:
Made some changes to the workflow to improve quality, see the "About this version" section for details.
This workflow implements SAM3 for character masking. It uses looping to process long videos, and while it can theoretically generate infinitely long videos, in practice it is limited by VRAM and processing time.
It supports both Wan Animate modes: character animation and character swap.
It has been tested on 1-minute, 480p videos using an RTX 4090 with 64GB of RAM, though it supports 360p, 720p and 1080p resolutions too. My tests suggest you can generate up to 20 seconds at 1080p with an RTX 4090 using a video batch size of 80 frames (if you have the patience and the time for it).
Commentaire généré par l’IA
Explication générée par l’IA à partir des détails de la source et de la configuration. Les suggestions sont clairement signalées.
Use the Wan2.2 Animate + SAM3 workflow to animate a character from a video reference or insert one into an existing video; it uses a character reference image and produces a video.
SAM3 provides character masking, while looping supports processing long videos.
The workflow supports Wan Animate’s character animation and character swap modes.
Character animation uses a full reference image together with a video reference.
Character swap inserts a character in place in an existing video using SAM3 segmentation.
Use yellow nodes for inputs and configuration; red nodes contain instructions and helpful notes.
The listed node packs include cg-use-everywhere, comfyui_controlnet_aux, comfyui_essentials, comfyui_layerstyle, comfyui_sam3, comfyui-custom-scripts, and comfyui-easy-use.
The listed node packs also include comfyui-impact-pack, comfyui-kjnodes, comfyui-longlook, comfyui-videohelpersuite, comfyui-wananimatepreprocess, and rgthree-comfy.
The listed model files are clip_vision_h.safetensors, dw-ll_ucoco_384_bs5.torchscript.pt, umt5_xxl_fp8_e4m3fn_scaled.safetensors, vitpose-l-wholebody.onnx, wan_2.1_vae.safetensors, Wan2_2-Animate-14B_fp8_e4m3fn_scaled_KJ.safetensors, yolov10m.onnx, and yolox_l.onnx.
For a face-swap attempt, set CHARACTER ISOLATION DESCRIPTION to "face".
For a separate face edit, use the edited face image as the Character Reference Image with character animation.
Suggestion · non vérifié
The description lists 360p, 720p, and 1080p as supported resolutions and reports a 480p example.
Before setup, verify the package sources for comfyui_sam3 and comfyui_controlnet_aux.
Suggestion · non vérifié
Infinite-length generation is theoretical, and practical video length is limited by VRAM and processing time; reported 1-minute 480p examples and suggested 20-second 1080p results with an 80-frame batch are planning references, not guarantees.
Suggestion · non vérifié
Faut-il modifier cela ?
Connectez-vous pour envoyer une demande de modification.
Sources
1 sourceExtraits de sources
1 extraitSource context: 1429 downloads · Type Workflows · Base model Wan Video 2.2 I2V-A14B
Features:
Character animation:
Full reference image animation using a video reference.
Character swap:
Insert a character in-place in a video using high quality segmentation with SAM3.
Face swap:
Possible using "face" as "CHARACTER ISOLATION DESCRIPTION", but don't expect miracles. You can achieve better results by face swamping the face separately using some edit model like Flux Klein or Qwen edit and then use that new edited image as your Character Reference Image using the character animation feature.
Detailed instructions are contained within the workflow itself:
Estimation des besoins VRAM
Estimation indisponible
24,8 GB sur 4 de 8 fichiers de modèle. Total des fichiers du modèle + 25 % de surcharge de chargement + 2 Go de tampon d'exécution, arrondi à l'unité supérieure.
Exigences
Exigences 21Vision model · 1.18 GB · Hugging Face · Comfy-Org/Wan_2.1_ComfyUI_repackaged
Sources possibles
dw-ll_ucoco_384_bs5.torchscript.pt
Non résoluCheckpoint · Unknown
Text encoder · 6.27 GB · Hugging Face · Comfy-Org/Wan_2.1_ComfyUI_repackaged
Sources possibles
vitpose-l-wholebody.onnx
Non résoluCheckpoint · Unknown
VAE · 242 MB · Hugging Face · Comfy-Org/Wan_2.2_ComfyUI_Repackaged
Checkpoint · 17.1 GB · SAFETENSORS · fp8 · full · Hugging Face · Kijai/WanVideo_comfy_fp8_scaled
yolov10m.onnx
Non résoluCheckpoint · Unknown
yolox_l.onnx
Non résoluCheckpoint · Unknown
Pack de nœud · Registry
comfyui_controlnet_aux
PossiblePack de nœud · Registry
Pack de nœud · Registry
Pack de nœud · Registry
comfyui_sam3
PossiblePack de nœud · Registry
Pack de nœud · Registry
Pack de nœud · Registry
Pack de nœud · Registry
Pack de nœud · Registry
Pack de nœud · Registry
Pack de nœud · Registry
Pack de nœud · Registry
Pack de nœud · Registry