My stuff

  • My Workflows

  • Liked Workflows

  • Following Workflows

Go to OpenArt main site
Upload workflow

Flux+InstandID Face-changing with multiple people

5.0

0 reviews
51
11.6K
2.4K
9
Description

该工作流主要分成三个板块:

1、Flux根据参考图生图;

2、SD放大工作流;

3、InstandID多人换脸;

操作步骤:

1、上传参考图,利用Flux进行生图类似的图,主要用到Florence2模型进行反推;

2、在换脸区,选择一张效果较好的图,并且上传图像中对应需要换脸的目标图。

The workflow is mainly divided into three sections:

1. Flux generates pictures based on reference pictures;

2. SD amplification workflow;

3. Face-changing for multiple people in InstandID;

Operation steps:

1. Upload the reference image and use Flux to generate similar images. The Florence2 model is mainly used for inference;

2. In the face-changing area, select a picture with better effect, and upload the image corresponding to the target image that needs to be changed.

Discussion

(No comments yet)

Loading...

Author

1
2.4K
51
11.6K
    微信图片_20241003024631.png (596.6 kB)

No reviews yet

  • - latest (a year ago)

Primitive Nodes (65)

Anything Everywhere (3)

CLIPTextEncodeFlux (1)

DownloadAndLoadFlorence2Model (1)

Fast Groups Muter (rgthree) (1)

Florence2Run (1)

GetNode (10)

Image Comparer (rgthree) (2)

JWStringConcat (2)

Note (8)

Pipe from any [Crystools] (2)

Pipe to/edit any [Crystools] (1)

Primitive string multiline [Crystools] (1)

Reroute (3)

Seed (rgthree) (1)

SetNode (9)

Switch any [Crystools] (1)

Switch image [Crystools] (1)

Switch latent [Crystools] (1)

easy forLoopEnd (1)

easy forLoopStart (1)

easy getNode (9)

easy setNode (5)

Custom Nodes (56)

ComfyUI

  • - BasicGuider (1)

  • - RandomNoise (1)

  • - BasicScheduler (1)

  • - KSamplerSelect (1)

  • - EmptyLatentImage (1)

  • - VAEEncode (1)

  • - SamplerCustomAdvanced (1)

  • - UNETLoader (1)

  • - VAELoader (1)

  • - DualCLIPLoader (1)

  • - CLIPTextEncode (1)

  • - VAEDecode (1)

  • - MaskToImage (1)

  • - ImageFromBatch (1)

  • - DifferentialDiffusion (1)

  • - InpaintModelConditioning (1)

  • - LoadImage (6)

  • - PreviewImage (4)

  • - LoraLoaderModelOnly (1)

  • - ControlNetLoader (1)

  • - UpscaleModelLoader (1)

ComfyUI Easy Use

  • - easy showAnything (1)

  • - easy imageCount (1)

  • - easy fullkSampler (1)

  • - easy pipeOut (1)

  • - easy a1111Loader (1)

  • - MaskFromBatch+ (1)

  • - ImpactSEGSToMaskBatch (1)

  • - SEGSPreview (1)

  • - ImpactSimpleDetectorSEGS (1)

  • - UltralyticsDetectorProvider (1)

  • - SAMLoader (1)

  • - LoadImagesFromDir //Inspire (1)

  • - ApplyInstantID (1)

  • - InstantIDModelLoader (1)

  • - InstantIDFaceAnalysis (1)

  • - LayerMask: MaskGrow (1)

  • - InpaintStitch (1)

  • - InpaintCrop (1)

  • - Int (2)

  • - ColorMatch (1)

  • - ImageBatchMulti (1)

  • - ImageResizeKJ (1)

  • - UltimateSDUpscale (1)

  • - CLIP_Interrogator (1)

Checkpoints (0)

LoRAs (1)

FLUX/F.1 Asian Girl亚洲人像_v1.0.safetensors