Go to OpenArt main site
Upload workflow

FLUX Redux + PULid Image to Image Face Swap Consistent Character

5.0

0 reviews
15
3.2K
909
7
Description

WATCH VIDEO TUTORIAL:

https://youtu.be/DLUHZUPBvEQ

What You'll Learn in This Tutorial:

  1. FLUX Redux Overview: An adapter for FLUX.1 base models, enabling refined image-to-image generation with controlled variations for restyling.
  2. FLUX Pul ID Explained: A method to generate images based on a single reference image, preserving key facial features while varying style, clothing, or background.
  3. Combining the Power of Both: Learn how FLUX Redux ensures visual diversity while FLUX Pul ID maintains face consistency for refined outputs.

Why Use This Workflow?

By merging FLUX Redux and FLUX PulID, this workflow is perfect for:

  • Artists and creators looking for consistent character designs.
  • Advertisers needing identity retention across campaigns.
  • Anyone wanting to experiment with advanced AI-driven face swaps.

Key Steps Covered:

  • How to download and install the necessary models:FLUX 1 Redux Dev ModelSIG CLIP Vision ModelPul ID Flux Model (new version)
  • Setting up ComfyUI and handling potential errors (e.g., the "transformer options" issue).
  • Step-by-step demonstrations using reference images like Taylor Swift, Elon Musk, Ana de Armas, and Gal Gadot.

Whether you're experimenting with male or female subjects, this workflow ensures your outputs are visually appealing and remarkably consistent.



Download Links and Resources:

  • Workflow: Available on Civit AI. Search "AI Motion Studio" for the latest version.
  • Models: Direct links provided in the video description.
  • Error Fixes: Detailed troubleshooting steps included for common issues.

Please Leave a Review on Civit AI: https://civitai.com/user/AIMotionStudio

Subscribe to our Youtube for ComfyUI Tutorials: http://www.youtube.com/@AiMotionStudio

Download CLIP Vision: https://huggingface.co/Comfy-Org/sigclip_vision_384/blob/main/sigclip_vision_patch14_384.safetensors

Download Style Model: https://civitai.com/models/969352/flux-adapter-flux1-redux-dev Download PuLID

Flux Model: https://huggingface.co/guozinan/PuLID/blob/main/pulid_flux_v0.9.1.safetensors

Download Workflow: https://civitai.com/models/989680/flux-redux-pulid-image-to-image-face-swap-consistent-character

Install Flux PuLID before you continue with this tutorial 👇🏻

Video Tutorial link: https://youtu.be/xUduNl7-pE0






Discussion

(No comments yet)

Loading...

Author

13
8.8K
123
29.5K
    ComfyUI_00096_.png (11.7 MB)
    ComfyUI_00098_.png (14.7 MB)
    ComfyUI_00097_.png (10.9 MB)

No reviews yet

  • - latest (a month ago)

Primitive Nodes (8)

ApplyPulidFlux (1)

FluxGuidance (1)

Image Comparer (rgthree) (1)

Note (1)

PulidFluxEvaClipLoader (1)

PulidFluxInsightFaceLoader (1)

PulidFluxModelLoader (1)

UnetLoaderGGUF (1)

Custom Nodes (19)

ComfyUI

  • - VAELoader (1)

  • - ImageScaleToTotalPixels (1)

  • - KSamplerSelect (1)

  • - BasicGuider (1)

  • - RandomNoise (1)

  • - SamplerCustomAdvanced (1)

  • - EmptyLatentImage (1)

  • - DualCLIPLoader (1)

  • - StyleModelApply (1)

  • - CLIPVisionEncode (1)

  • - CLIPTextEncode (1)

  • - VAEDecode (1)

  • - SaveImage (1)

  • - CLIPVisionLoader (1)

  • - StyleModelLoader (1)

  • - BasicScheduler (1)

  • - LoadImage (2)

KJNodes for ComfyUI

  • - GetImageSizeAndCount (1)

Checkpoints (0)

LoRAs (0)