Documentation
  • How Our System works
  • Text to Image
    • Dalle-3
    • SDXL Models
  • Image Editing
    • Replace Background
    • Inpainting and CleanUp
    • Upscaling
  • Image to Image
    • SDXL Models
    • Face Consistent
    • ControlNet
    • Illusion Diffusion
    • Image Variation
  • ChatBot
  • Knowledge Bases in Chatbot
  • Text To Speech
  • History Section
  • Qreative wall
  • How to Cancel Your Subscription
  • Extra Resources
    • Prompt Advancement
    • Private Session
    • Image Editing Options
    • Resolution
    • Generation Settings
  • API Platform
    • Text to Image
    • Image to Image
    • ControlNet
    • Inpainting
    • Replace Background
    • Face Consistency
    • Face Avatar
    • Image Variation
    • Illusion Diffusion
    • Upscaling
    • Background Removal
    • Text to Speech
    • Task Status
    • ChatBot API
Powered by GitBook
On this page
  • Selecting the Face Consistent Model
  • Providing the Prompt and Uploading an Image
  • Configuring the Image Generation
  • Adjusting the Strength Parameter
  • Generating the Images
  • Resetting and Editing the Images
  • Protecting Your Private Content

Was this helpful?

  1. Image to Image

Face Consistent

PreviousSDXL ModelsNextControlNet

Last updated 1 year ago

Was this helpful?

Qolaba AI offers a specialized "Face Consistent" model, which is a community-developed version of Stable Diffusion. This model is designed to generate images with a consistent face, making it ideal for scenarios where you need to create images of a specific person in different environments or settings.

Selecting the Face Consistent Model

To access the Face Consistent model, navigate to the "Model" section on the left panel and select the "Face Consistent" option.

Providing the Prompt and Uploading an Image

After selecting the Face Consistent model, you can enter your prompt or text description in the prompt text box. For more information on crafting effective prompts, including the use of magic prompts, style presets, negative prompts, and keywords, please refer to the Prompt Advancement.

Additionally, you can upload an image from the left panel, which the model will use as a reference for the face consistency.

Configuring the Image Generation

On the right panel, you can select the number of images you want to generate based on your prompt.

Below this, you'll find options to choose the desired aspect ratio, such as ultrawide, portrait, or square. For more details on resolution, please refer to the Resolution Guide.

Adjusting the Strength Parameter

The Face Consistent model also includes a "Strength" parameter, which allows you to control the intensity of the face generation. A higher value will lead to a more exact match with the reference face, while a lower value may result in a slightly different face. However, this parameter may not have a significant impact in every case.

For more information on the prompt guidance, please refer to the Generation Settings Guide.

In addition to the other settings you can configure, Qolaba AI also provides an option to select the inference type for image generation. This setting can be found on the left panel, right next to the "Generate" button. You can choose from three inference types: Normal, Fast, and Hyper. The Normal option is cost-efficient and fast, while the Fast and Hyper options can generate images more quickly but at a higher cost. For more details, please refer to the Generation Settings Guide.

Generating the Images

Once you've configured all the desired settings, click the "Generate" button on the left panel to start the image generation process. The Face Consistent model will then generate the images based on your input.

Resetting and Editing the Images

If you need to start over or make changes to your settings, you can find a "Reset" option at the bottom of the right panel.

After generating the images, you can access the image editing tools by clicking the hamburger menu. For more details on the available editing features, please consult the Image Editing Guide.

Protecting Your Private Content

Qolaba AI also offers a "Private Content" option, which allows you to keep your generated images secure and hidden from others. For more information on this feature, refer to the Private Session Guide.

Original Image (AI generated face)
Image from face consistent model (Prompt : amazing cute woman in market purchasing vegetables, Style Preset : Photographic)