TABLE OF CONTENTS
Overview
Control Adapters are a powerful set of tools enabling you to control and guide the image generation process to better produce a desired outcome. Utilizing Control Adapters is a key piece to success with AI image generation.
There are three types of Control Adapters that are supported by Invoke:
- ControlNet
- Image Prompt Adapter (IP-Adapter)
- Text to Image Adapter T2I-Adapter
ControlNet and T2I-Adapter are similar, with many different types of models offered, each suited to a different purpose. Some examples of popular models are Depth, Canny, and OpenPose.
To learn more about using the different models offered & how to use ControlNet and T2I-Adapters, take a look at the articles below:
IP-Adapter uses the concepts that the AI can detect in your imagery to inspire new content. To learn more about IP-Adapter, see this article:
Settings
There are some settings that apply to all three types of Control Adapters. Additionally, each Control Adapter section can be expanded in order to manipulate settings for the image pre-processor that adjusts your uploaded image before using it when you Invoke.
Weight
The strength of the Control Adapter. Higher weights will follow the guidance of the Control Adapter more closely.
Begin / End Step Percentage
What portion of the generation process that the Control Adapter will be applied during. A value of 0 represents the start of the generation, and a value of 1 represents the end. Control Adapters applied during the start of the generation generally help guide structure and composition, while Control Adapters applied towards the end of the generation will affect details in the image.
Was this article helpful?
That’s Great!
Thank you for your feedback
Sorry! We couldn't be helpful
Thank you for your feedback
Feedback sent
We appreciate your effort and will try to fix the article