Image 2 Image
Stable diffusion can also take existing images as input, like your brokers!
Last updated
Stable diffusion can also take existing images as input, like your brokers!
Last updated
To use this feature, use the Advanced Prompting !dream command like you normally would and simply attach an image in Discord before submitting the message.
NOTE: If you are using CyberBrokers as input, make sure to use one with a background. Or said the opposite, don't use one with a transparent background.
There is an additional option you will want to use with this feature: --strength.
This option takes a value from 0.1 to 0.9 and controls how closely stable diffusion will keep to the input image. The closer to 0.1 you go, the more original the output will be. The closer to 0.9 you go, the more freedom AI has and the more the prompt you provide will direct output.
For the prompt, describe what is in the image so stable diffusion can help reproduce it.
For finer levels of control use the --npa true option to add in your own positive/negative prompts.
!dream male, front view, wavy flowing white hair, sun glasses, 3d render --strength 0.5
Give it a try with your broker or any image!
