This project integrates segmentation, editing, and generation processes. Using the Segment Anything Model (SAM), we perform image segmentation, select specific regions for replacement, and employ a stable diffusion inpainting model from Hugging Face to generate new images based on prompts. This versatile approach allows for applications like clothes editing. Additionally, an alternative version of the project involves utilizing the Clipseg model for image segmentation based on text prompts.
Note: this repository and its README will be completed over time.