AI Product Photographer

How to set it up?

• Clone this repo
• Install the dependencies of the requirements.txt file
• Download the Segment Anything Model

How to use it?

• Run with streamlit run main.py (it takes some time at the first run to download the Stable Diffusion images from Hugging Face)
• Upload your product picture

• Click on the product for helping the segmentation and generate the mask

• Once the mask has been generated, write a prompt for the AI to generate the inpainting

• Obtain the final result (and iterate with the prompt if needed)