💡 How to Use This Demo:

  1. Upload an identity (ID) refernces. The model will generate images containing the same identities. You can upload 1-4 reference images, each containing one clear human face.
  2. Provide detailed prompts describing the identity. Please refer to Tips under the Generated Image field. WithAnyone is "controllable", so it needs more information to be controlled.
  3. [Recommended] Specify bounding boxes for each identity. You can either:
    • Upload a multi-person image containing all the desired identities, and the system will automatically detect faces and extract bounding boxes. (Recommended for most users)
    • Manually input bounding boxes in the text box (format: x1,y1,x2,y2, one per line). If left empty, the system will automatically generate bounding boxes based on the number of reference images.
  4. [Recommended] Adjust the "Spiritual Resemblance <--> Formal Resemblance" slider to balance between preserving identity details and allowing creative freedom. Move it to the right for more resemblance to the reference images, or to the left for more creativity.
  5. Click the "Generate" button to generate an image. Enjoy!
0 1
512 1024
512 1024
1 50
1 10

Face References (1-4 required)

Mask Configuration (Option 1: Automatic)

Mask Configuration (Option 2: Manual)

Example Configurations

Tips for Better Results

Be prepared for the first few runs as it may not be very satisfying.

  • Provide detailed prompts describing the identity. WithAnyone is "controllable", so it needs more information to be controlled. Here are something that might go wrong if not specified:
    • Skin color (generally the race is fine, but for asain descent, if not specified, it may generate darker skin tone);
    • Age (e.g., intead of "a man", try "a young man". If not specified, it may generate an older figure);
    • Body build;
    • Hairstyle;
    • Accessories (glasses, hats, earrings, etc.);
    • Makeup
  • Use the slider to balance between "Resemblance in Spirit" and "Resemblance in Form" according to your needs. If you want to preserve more details in the reference image, move the slider to the right; if you want more freedom and creativity, move it to the left.
  • Try it with LoRAs from community. They are usually fantastic.
Click to load example configurations
Prompt Generation Width Generation Height Guidance Number of steps Seed (-1 for random) Reference 1 Reference 2 Manual Bounding Boxes (one per line, format: x1,y1,x2,y2) Spiritual Resemblance <--> Formal Resemblance