README.md

metadata

title: AI Age Progression Studio
emoji: ⏳
colorFrom: green
colorTo: blue
sdk: streamlit
sdk_version: 1.46.1
app_file: app.py
pinned: false

AI Age Progression Studio

This Hugging Face Space hosts an AI-powered application that transforms faces in photos to different ages. You can upload an image and choose from various transformation methods:

• Transform to Specific Age: Set a precise target age.
• Transform with Custom Prompt: Use a descriptive text prompt to guide the age transformation (e.g., "a photo of an 80-year-old woman with wrinkles").
• Transform to Current Age (Dynamic): Calculate the current age of the person in the photo based on the photo's taken date and their age at that time.

The application uses a Stable Diffusion model with a specialized LoRA (Low-Rank Adaptation) for age progression, along with a face alignment model to ensure optimal results.

How to Use

1. Upload Your Photo: Use the file uploader in the sidebar to select a clear photo of a face.
2. Choose Transformation Method: Select one of the three options in the sidebar.
3. Adjust Settings: Depending on your chosen method, configure the target age, custom prompt, or date/age details. You can also adjust the "Transformation Strength."
4. Generate: Click the "✨ Generate Transformed Image ✨" button.
5. View & Download: The transformed image will appear on the right, and you'll have an option to download it.

Important Notes & Limitations

• The AI model may struggle with very young (1-5 years) or very old (80-90+ years) transformations.
• Significant transformations (high 'Strength') can alter facial features, making the generated person look less like the original.
• For best results, use clear, well-lit photos with a single, prominent face.

Technical Details

• Framework: Streamlit
• AI Models:
  • runwayml/stable-diffusion-v1-5 (Base Diffusion Model)
  • navmesh/Lora (LoRA for Age Progression)
  • face-alignment (for face detection and alignment)

Deployment

This application is designed for deployment on Hugging Face Spaces. The requirements.txt file lists all necessary Python dependencies. ```

Key additions in the YAML front matter:

• title: The title that will appear for your Space.
• emoji: An emoji to represent your Space.
• colorFrom, colorTo: Used for the gradient background on the Space card.
• sdk: Specifies the SDK used (e.g., streamlit, gradio, docker).
• sdk_version: The version of the SDK. I've put 1.36.0 as a common recent version, but you can check the Streamlit documentation for the absolute latest or a version you prefer.
• app_file: Points to your main application file, which is app.py.
• pinned: Whether the Space should be pinned to your profile.

Make sure to replace your existing README.md with this updated version when deploying to Hugging Face Spaces.