GithubHelp home page GithubHelp logo

Comments (5)

cloneofsimo avatar cloneofsimo commented on May 21, 2024 1

Say you have a input_prompt = "a rabbit and a cat".
and you have text-labeled segmentation maps with region with dog in it. In this current implementation, this will cause an error because dog is not in the input_prompt.
Make sure that that all your segmentation labels are strictly include in the prompt.

Note : This is, to be honest, not the best way to implement as there can always be subtle difference in segmentation label and prompt, which is reasonable! I am planning to thus implement some sort of "neural" version of this, but this was not in their paper so I don't know if this would work.

from paint-with-words-sd.

hiss-remi avatar hiss-remi commented on May 21, 2024 1

Say you have a input_prompt = "a rabbit and a cat". and you have text-labeled segmentation maps with region with dog in it. In this current implementation, this will cause an error because dog is not in the input_prompt. Make sure that that all your segmentation labels are strictly include in the prompt.

The thing that was tripping me up is that if you have multiple words for a given region, it won't trigger this error as long as one of the words is in your prompt. Unfortunately, words that aren't in the prompt usually don't seem to have any effect on the output image, while words that are in the prompt that could apply to the subjects of multiple regions frequently end up contaminating them. Which makes using this for complicated subjects a bit of a crapshoot.

(Although the lack of negative prompting is sometimes making it hard to tell what's going on... the model I'm using is very good at making bad art if you don't tell it not to.)

from paint-with-words-sd.

cloneofsimo avatar cloneofsimo commented on May 21, 2024 1

could apply to the subjects of multiple regions frequently end up contaminating them. Which makes using this for complicated subjects a bit of a crapshoot.

Right, I am aware of this issue. This is fixable problem and I will work on it.

(Although the lack of negative prompting is sometimes making it hard to tell what's going on... the model I'm using is very good at making bad art if you don't tell it not to.)

I have added negative prompting today, so check it out!
you can use it with arguement unconditional_input_prompt: str = "",

from paint-with-words-sd.

leonnn1 avatar leonnn1 commented on May 21, 2024

thank you, I mannaged to make it work, I am still a bit lost but it's doing something haha, great work by the way!

from paint-with-words-sd.

hiss-remi avatar hiss-remi commented on May 21, 2024

Right, I am aware of this issue. This is fixable problem and I will work on it.

Glad to hear it. I... don't really get how any of this works, so I wasn't sure what I should hope for.

from paint-with-words-sd.

Related Issues (20)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.