[AVD] SUPERFAST Text to Video with Clothing and Face Swap
5.0
0 reviewsDescription
Workflow for Advanced Visual Design class.
Note: This requires KJNodes (not in comfymanager) for the GET and SET nodes: https://github.com/kijai/ComfyUI-KJNodes
You can install it in the ComfyManager by going to "Install via Git URL" in the options in the Manager. Otherwise may need to do it manually.
Instructions:
- You will need to replace the image loaders with your own images
- Text to Image Group (three of them)
- FACEID Group
- You can easily bypass any group (for example disabling the second render pass until ready) using the controls in the top left. This also serves as an outline for the order of all the groups.
- This workflow at its core is optimized for using LCM rendering to go from text to video quickly. To help with its stylistic direction, you can use IPAdapters to control both style and face of the video output.
- For this example we use IPAdapters to control the clothing and the face.
- This one has pretty thorough comments so I won't say too much here about it, although playing with the weights of the IP Adapters will change how they influence the output render in addition to the text (which comes in two prompts here: one for overall style, and a second for minor details).
- The post processing flow uses a second rendering pass and a frame interpolator for smoothing, upscaling, and re-rendering the output.
Discussion
(No comments yet)
Loading...
Resources (2)
Reviews
No reviews yet
Versions (1)
- latest (2 years ago)
Node Details
Primitive Nodes (74)
Fast Groups Muter (rgthree) (1)
GetNode (31)
Note (21)
Primitive boolean [Crystools] (1)
Primitive integer [Crystools] (1)
Reroute (2)
SetNode (16)
Switch image [Crystools] (1)
Custom Nodes (62)
- ADE_StandardUniformContextOptions (2)
- ADE_AnimateDiffSamplingSettings (2)
- ADE_UseEvolvedSampling (2)
- ADE_ApplyAnimateDiffModel (2)
- ADE_LoadAnimateDiffModel (2)
- ADE_AnimateDiffLoRALoader (2)
ComfyUI
- ModelSamplingDiscrete (1)
- LoraLoaderModelOnly (1)
- CLIPVisionLoader (1)
- EmptyLatentImage (2)
- KSampler (1)
- VAEDecode (3)
- SaveImage (1)
- CLIPTextEncode (4)
- CheckpointLoaderSimple (1)
- LoraLoader (2)
- LoadImage (5)
- ConditioningConcat (2)
- VAEEncode (1)
- UpscaleModelLoader (1)
- ImageUpscaleWithModel (1)
- FreeU_V2 (1)
- KSamplerAdvanced (2)
- ImageResize+ (2)
- RIFE VFI (1)
- ImpactInt (4)
- ImpactSwitch (1)
- PrepImageForClipVision (2)
- IPAdapterModelLoader (2)
- IPAdapterAdvanced (1)
- IPAdapterFaceID (1)
- VHS_VideoCombine (2)
- VHS_SelectEveryNthLatent (1)
- PlaySound|pysssss (1)
- ShowText|pysssss (1)
- Random Number (1)
Model Details
Checkpoints (1)
SD15_3D_3dAnimationDiffusion_v10.safetensors
LoRAs (5)
Cyberpunk _Anime_sdxl.safetensors
LCM_LoRA_Weights_SD15.safetensors
Range_murata_style_v02.safetensors
v2_lora_ZoomIn.ckpt