Innovative CAD generator using Stable Diffusion
Uses the newly published Stable Diffusion model to generate CAD models/blueprints.
For image recognition/classification the model tensorflow is used.
For UI the module tkinter is used.
pip3 install -r requirements.txt
python3 main.py
You must have a good internet connection since an api is being called.
You must provide a CAD model or a blueprint otherwise it will generate incorrect results.
If it does not generate the desired results make sure the images are in RGB mode.(optional)
Use the example.png as a test image to generate sample output.
If it gives an error during installation make sure your Python version is before 3.11 since the new one is conflicting the packages.
Tested on Python 3.10/3.10.8. If tensorflow throws an AVX instruction error that means that your CPU is not supported. Please check if your CPU supports AVX https://10scopes.com/cpu-support-avx/.
If an error occurs while generating the image after the recognition, please check if you added your api key on line 16.
To get a new API key please visit https://beta.dreamstudio.ai/membership .
https://numpy.org/ (used for expanding the dimensions of the image)
https://www.tensorflow.org/ (image classification)
https://beta.dreamstudio.ai/ (the api that generates the new image, limited to about 124)
https://docs.python.org/3/library/tkinter.html (the library used for the UI)
https://pillow.readthedocs.io/ (used for displaying the images)