AI Rug Placement Web API Development

We are looking to develop a Python-based Web API that takes a room photo and a rug design image, and uses AI-powered image processing techniques to realistically place the rug onto the floor in the room.

The system should be accessible via HTTP (REST API), accept file uploads, and return a composite image with the rug seamlessly placed on the floor. This service will later be integrated into our project.

Key Requirements:
- Must be developed in Python
- HTTP (REST API) accessible
- A single endpoint (e.g., POST /generate) should:
- Accept two uploaded files: room.jpg (room photo) and rug.png (rug photo)
- Process the images and realistically place the rug at the center of the floor
- Return the final composited image (image/png, base64)
- AI-powered segmentation and/or image processing methods must be used

AI Usage:

- Floor segmentation and rug positioning must use AI-based techniques, such as:
- Segment Anything (SAM), U-2-Net, DeepLab, Mask R-CNN, etc
- Rug warping and blending can be implemented via OpenCV homography, AI-based surface projection, or similar methods
- The final image must look photorealistic, with the rug naturally integrated into the floor.
No modifications, alterations, or enhancements should be made to the original room photo or the rug design itself only realistic placement is required.

!Important;
- No manual masking.

Deliverables:

- Fully working Python project
- requirements.txt and README.md with setup instructions
- Example input images and result (at least three sample pair with different image)
- Simple API documentation (endpoints, request/response format)

Ideal Skills and Experience:
- Proficiency in Python and web API development
- Experience with AI and image processing techniques
- Familiarity with handling image uploads and returning processed images

This service will be integrated into a project, so clean, modular code and thorough documentation are essential.