Object Detection and Replacement Project

Overview

This project utilizes AI models to detect objects in images and replace them with new content. The system performs the following tasks:

• Detects objects using a zero-shot detection model (Grounding DINO).
• Generates segmentation masks using Segment Anything (SAM).
• Replaces the detected objects with new content via Stable Diffusion XL inpainting, driven by text prompts.

Requirements

• Python 3.x
• Libraries: torch, transformers, diffusers, cv2, PIL, requests, numpy

Installation

1. Clone the repository.
2. Install the dependencies using:

   pip install diffusers transformers accelerate torch -U

3. Login to HuggingFace Hub:

   from huggingface_hub import login
   login(token="your_huggingface_token")

4. Use the script to process images.

Usage

Run the script and provide an image URL and desired labels. The system will detect objects, generate masks, and replace the objects in the image based on the prompt.

Result

License

This project is licensed under the MIT License. See the LICENSE file for more details.

---