AI Image Enhancement & Retouching

I want to build an AI-powered solution that automatically enhances and retouches images. The core idea is to feed photographs into the system and receive back professionally polished results—improved sharpness, balanced exposure, refined skin texture, and overall visual pop—without manual intervention each time.

Here’s what I need from you:
• Design or adapt a suitable model (GAN, diffusion, or other state-of-the-art architecture you trust) focused on image enhancement and retouching.
• Implement the full pipeline—pre-processing, inference, and post-processing—so I can drop in a batch of images and retrieve the retouched versions in a predictable folder structure.
• Keep quality high: no lost details, no over-smoothing, colors must remain natural.
• Provide clean, well-commented code (Python preferred, using libraries such as PyTorch, TensorFlow, or OpenCV).
• Include a brief README describing environment setup and how to run the script or notebook.

Acceptance criteria:
1. Given a sample test set I supply, the AI consistently improves perceived image quality (brightness, contrast, clarity) and removes minor blemishes without introducing artifacts.
2. Processing time per 4K photo should not exceed 10 seconds on a single modern GPU.
3. Deliverables compile and execute without errors on a fresh environment following your setup guide.

If you have prior experience with AI image editors or have demo links, feel free to share them—seeing before/after examples will help me gauge fit quickly.