Skip to main content
Participant
March 24, 2023
Released

P: Negative prompting

  • March 24, 2023
  • 70 replies
  • 31054 views

In the text to image part, I didn't find a place to cancel prompts like in stable diffusion. Often we have artifacts with double faces or eyes in the wrong places. It depends on the channels I think but it could be efficient and necessary to add this function. Or maybe add a basic lexicon for prompts if we have to use No in front of cancels or others basic operations.

 

Thank you for your job, im a teacher and i think you are the best to create an democratic useful app for commons users

 

Eric

70 replies

Known Participant
July 2, 2023

I don't get it. [avoid=xxx] seems to do the opposite of what is expected - it does the same as typing "xxx" without the [avoid=] part, i.e. like typing it in as regular positive prompt. Or does this concern Firefly only, not PS Generative Fill?

Participant
June 30, 2023

Can you use a minus sign or similar to avoid or negate certain words when generating images with Firefly AI?

 

For example, I want a plain skull but keep getting Day of the Dead sugar skulls (flowers and intricate designs)--Is there or could there be a way to match "skull" but not "sugar skull"?

Participant
June 20, 2023

Through the many AI Art programs that I have tried, there was one in particular that stood out to me because it had a feature called "negative prompt". This was where you would enter what you did not want to be in your picture. For example, if I wanted to generate an image but didn't want text, I would put in the negative prompt, "text". This can also be used for blurry, specific items, certain colors, etc. You would still put in your normal prompt as usual but the negative prompt would be below just in case.

Participant
June 5, 2023

Add a negative prompts, and Let us chose number of images generated... 2 images instead of 4 etc.

 

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

J E L
Community Expert
Community Expert
June 5, 2023

Hi @Akila30302169eqiv, for a negative prompt, you can try using the Firefly code: [avoid=] as in [avoid=green, cars]. Let us know if that helps. The idea of choosing the number of generated images is a good suggestion.

Anitaphotography.nl
Participating Frequently
June 2, 2023

The [avoid = XXX] does absolutely nothing.
There needs to be a separate text box to list the unwanted content.

Participant
June 1, 2023

A feature that is the opposite of the normal promp for what to include in the picture. Most other AI image generators like Leonardo.ai have this feature. I believe it would make it more effient in making the correct images.

 

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

Participant
May 31, 2023

Windowless room, room without windows and shabby room, etc all give rooms with windows. No negative prompt either so no way to generate something like this which should easily be possible.

 

This is just an example, but refining anything to remove an undesirable element doesn't work. For example, no cleavage.

 

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

Participant
May 15, 2023

I need a negative promt to  take out things that I don't want to be in the image

 

Merging your thread here so we can track the demand for certain features.

Participant
May 15, 2023

????

Participant
May 14, 2023

Hello Adobe Firefly,

I would like to start by saying how much I am a fan of your image generation platform. However, I would like to suggest two improvements that could make the user experience even better.

Firstly, the implementation of a NEGATIVE PROMPT would be a great addition to the platform. As many users may agree, the option of a negative prompt would allow users to specify which elements they do NOT want in their images, which would save a lot of time and increase the accuracy of the results. Imagine being able to tell the platform that you do not want a specific object to appear in your image, instead of having to specify all the objects you want to include. This would be a valuable addition to the platform.

Additionally, another suggestion would be the implementation of ready-made templates and styles for image generation. With ready-made templates and styles, users could generate images quickly and easily, without having to go through the process of customizing every detail. This would make the platform more accessible for beginner users, as well as saving time and increasing the productivity of more experienced users.

Thank you for considering these suggestions, and I look forward to exciting future updates on the platform.

Sincerely, Felipe

Ares Hovhannesyan
Community Expert
Community Expert
May 14, 2023

I catch that in Stable Deffusion but never use it. 

Anitaphotography.nl
Participating Frequently
May 4, 2023

The [avoid=XXX] doesn't fully seem to get the job done though...