Wan 2.1 AI Talking Avatar. (Vace, chatterbox, float) V2
5.0
0 reviewsDescription
This is V2 of my talking Avatar flow, V3 is coming soon with full extending for much longer videos.
Example of what is coming in v3 - https://www.youtube.com/shorts/GsDXeiWx6n0
This uses 3 main concepts to create talking Avatar videos.
Chatterbox
- Text to speech with audio changing options.
Float Lip Sync
- Generates a head face sync video.
Vace outpaint
- Generates a video outpainted from the face sync based on your ref image
- This can use vace 1.3 or 14b models for the outpaint.
Many notes included.
V3 in a few days :)
Discussion
(No comments yet)
Loading...
Reviews
No reviews yet
Versions (1)
- latest (6 months ago)
Node Details
Primitive Nodes (135)
Any Switch (rgthree) (3)
Audio Sequence (mtb) (5)
Audio Stack (mtb) (1)
Display Any (rgthree) (1)
Display Int (rgthree) (4)
FL_ChatterboxTTS (3)
FL_ChatterboxVC (1)
Fast Groups Bypasser (rgthree) (19)
FloatProcess (1)
GetNode (19)
Image Batch To Sublist (mtb) (4)
Image Comparer (rgthree) (1)
ImagePadKJ (1)
ImageResizeKJv2 (6)
LoadFloatModels (1)
Lora Loader Stack (rgthree) (1)
MaskFix+ (3)
MaskPreview (1)
Note (19)
PrimitiveNode (18)
ReActorFaceBoost (1)
ReActorUnload (3)
Reroute (4)
SetNode (10)
TrimVideoLatent (1)
UnetLoaderGGUF (1)
UnetLoaderGGUFDisTorchMultiGPU (1)
VocalAndSoundRemoverNode (1)
WanVaceToVideo (1)
Custom Nodes (91)
- PreviewAudio (12)
- CR Model Input Switch (2)
- CR Image Input Switch (1)
ComfyUI
- ImageFromBatch (4)
- MaskToImage (2)
- VAELoader (1)
- CLIPLoader (1)
- LoraLoaderModelOnly (3)
- UNETLoader (1)
- ImageScaleBy (1)
- UpscaleModelLoader (1)
- PreviewImage (2)
- JoinImageWithAlpha (1)
- VAEDecode (1)
- CLIPTextEncode (2)
- LoadImage (1)
- KSamplerAdvanced (1)
- GetImageSize+ (1)
- BatchCount+ (1)
- SimpleMath+ (1)
- ImpactSimpleDetectorSEGS (3)
- SegsToCombinedMask (3)
- ImageListToImageBatch (2)
- MaskListToMaskBatch (2)
- ImpactGaussianBlurMask (1)
- UltralyticsDetectorProvider (1)
- LoadAudio (1)
- VHS_VideoCombine (5)
- VHS_LoadAudioUpload (7)
- ColorToMask (1)
- ImageAndMaskPreview (1)
- INTConstant (3)
- ImageUpscaleWithModelBatched (1)
- GetImageSizeAndCount (2)
- VRAM_Debug (1)
- Change Channel Count (6)
- Mask To Region (1)
- Cut By Mask (1)
- Paste By Mask (1)
- ReActorFaceSwapOpt (2)
- ReActorOptions (1)
- ReActorRestoreFace (2)
- Mask Invert (3)
Model Details
Checkpoints (0)
LoRAs (3)
vace\Wan21_CausVid_bidirect2_T2V_1_3B_lora_rank32.safetensors
wan\control\Wan21_AccVid_T2V_14B_lora_rank32_fp16.safetensors
wan\control\Wan21_T2V_14B_MoviiGen_lora_rank32_fp16.safetensors