✨ Prism AI Studio ✨
🪄 Effortless AI Upscaling 🪄

Get ready to witness some AI magic! 🚀 Prism AI Studio is a super-sleek, easy-to-use web application that lets you upscale your images to stunning high resolutions using the power of Real-ESRGAN and optionally enhance faces with Prism-AI-Studio. All wrapped up in a vibrant, interactive interface built with FastAPI and vanilla JavaScript! 🎨

---

🌟 Key Features 🌟

• ✨ AI-Powered Upscaling: Leverages state-of-the-art Real-ESRGAN models to intelligently enlarge your images.
• 😊 Optional Face Enhancement: Integrates Prism-AI-Studio to magically restore and improve faces in your upscaled images (if Prism-AI-Studio is installed).
• 🖥️ Sleek Web Interface: Modern UI built with FastAPI (backend) and pure HTML/CSS/JavaScript (frontend). No heavy frameworks needed!
• 🔮 Glassmorphism UI & Themes: Features a cool glassmorphism design with multiple vibrant themes (Vibrant, Pastel Dream, Ocean Breeze, Sunset Glow) to suit your mood! 🎨
• 🧠 Multiple Model Support: Choose from various Real-ESRGAN models (General, Anime, etc.) tailored for different image types.
• ☁️ Automatic Model Downloads: Required AI models are downloaded automatically on first use, with fancy progress updates!
• 📊 Real-time Progress: Stay informed with WebSocket-powered status updates for uploads, downloads, and the enhancement process.
• 🖱️ Drag & Drop Uploads: Easily upload multiple images by dragging them onto the app.
• 🖼️ Instant Previews: See thumbnails of your input images and the enhanced results as they're processed.
• ⚙️ Configurable Options: Fine-tune the upscaling factor, output format (PNG, JPG, WEBP), denoise strength (for specific models), and advanced tiling parameters.
• 💻 GPU & CPU Support: Automatically detects NVIDIA GPUs (via CUDA) for speedy processing, but gracefully falls back to CPU if needed. Option to force CPU usage.
• 💾 Download Options: Download individual enhanced images or grab them all at once in a convenient ZIP archive.
• 🗑️ Task Management: Remove individual images or clear the entire session, with cancellation support for ongoing tasks.
• 📱 Responsive(ish) Design: The UI adapts reasonably well to different screen sizes.

---

🛠️ Technology Stack 🛠️

• Backend: FastAPI, Uvicorn, WebSockets
• AI Models: Real-ESRGAN, Prism-AI-Studio (Optional)
• Image Processing: PyTorch, OpenCV-Python, basicsr
• Frontend: HTML5, CSS3, Vanilla JavaScript
• Core Libraries: requests, tqdm

---

🚀 Installation & Setup 🚀

Get ready to launch Prism AI Studio! Follow these steps:

Prerequisites:

• 🐉 Anaconda or Miniconda
• 🐍 Python 3.7 or Higher
• 📦 pip (Python Package Installer)
• ♨️ PyTorch >= 1.7
• ➕ Git Installed
• ❗NVIDIA GPU + CUDA (Optional)
• 🐧Linux (Optional)

1. Clone the Repository

git clone https://github.com/Md-Siam-Mia-Man/Prism-AI-Studio.git
cd Prism-AI-Studio

2. Create a Virtual Environment

conda create -n Prism-AI-Studio python=3.7 -y
conda activate Prism-AI-Studio

3. Install PyTorch

# Try to make sure your drivers are updated
# For NVIDIA GPU
conda install pytorch torchvision torchaudio pytorch-cuda=<your_cuda_version> -c pytorch -c nvidia -y

# For CPU
conda install pytorch torchvision torchaudio cpuonly -c pytorch -y

4. Install Dependencies

pip install -r requirements.txt

💻 Usage

▶️ Running the Application

Start the Flask Server:

python prism.py

For one-click RUN Create a new Prism-AI-Studio.bat Batch File on your Prism-AI-Studio directory using the script given below:

@echo off

:: Activate the conda environment for Prism-AI-Studio
CALL "C:\ProgramData\<Your Anaconda distribution name>\Scripts\activate.bat" Prism-AI-Studio

:: Navigate to the Prism-AI-Studio directory (Change path according to yours)
cd /D path/to/your/Prism-AI-Studio

:: Run the Prism-AI-Studio web interface script
python prism.py

Access the Web Interface:
🌐 Open your browser and visit: http://127.0.0.1:3005

Note: Check the exact versions required by the specific versions of Real-ESRGAN and Prism-AI-Studio you intend to use.

---

✨ How to Use ✨

Using Prism AI Studio is a breeze!

1. Access the App: Open http://127.0.0.1:3025 in your browser.
2. Upload Images: Drag and drop your image files onto the designated area, or click it to browse your files. Accepted formats: PNG, JPG/JPEG, WEBP, BMP, TIFF.
3. Configure Settings (Sidebar):
   • Choose your desired Upscaling Model.
   • Select the Upscale Factor (e.g., 2x, 4x).
   • Pick an Output Format (PNG recommended for quality).
   • Toggle Face Enhancement (Prism-AI-Studio) On/Off (if available).
   • Adjust Advanced Settings like Tile Size if needed (especially if you encounter memory errors).
   • Select a UI Theme if you like! 🌈
4. Model Downloads: If you're using a model for the first time, the app will automatically download it. You'll see progress notifications! ☁️➡️💻
5. Enhance! Once your images are uploaded and settings are configured, smash that "Enhance Images" button! ✨
6. Watch the Progress: See the status updates on the image thumbnails in real-time.
7. Download Results: Once completed, download buttons will appear on the enhanced image thumbnails. You can also use the "Download All" button to get a ZIP file of all successful enhancements. 📦

---

🧠 Models & Configuration 🧠

• Real-ESRGAN Models: The app supports several pre-defined models, downloaded automatically to the ./weights/ directory:
  • RealESRGAN_x4plus (Default, good general-purpose 4x)
  • RealESRNet_x4plus (Faster alternative to x4plus)
  • RealESRGAN_x4plus_anime_6B (Optimized for 4x Anime upscaling)
  • RealESRGAN_x2plus (General-purpose 2x)
  • realesr-animevideov3 (For Anime Video frames, 4x)
  • realesr-general-x4v3 (General-purpose 4x with denoise option)

---

🎨 Frontend Themes 🎨

Spice up your upscaling experience! Choose from several themes in the top-right dropdown:

1. Vibrant Glassmorphism (Default): The sleek, modern look with purple/blue gradients.
2. Pastel Dream: A softer, lighter theme with pastel orange/yellow tones.
3. Ocean Breeze: Cool and refreshing deep blue to cyan gradients.
4. Sunset Glow: Warm pink and orange hues for a cozy feel.

---

❗ Troubleshooting & Notes ❗

• CUDA Out of Memory Errors: If you see errors related to GPU memory during enhancement (especially with large images or high scale factors):
  • Try reducing the Tile Size in the "Advanced Settings". Start with values like 256 or 128. A value of 0 attempts to process the whole image at once.
  • Try reducing the Upscale Factor.
• CPU Performance: Processing on the CPU will be significantly slower than using a compatible NVIDIA GPU. Be patient!
• Model Downloads: Initial model downloads require an active internet connection and can take some time depending on your network speed and the model size. Progress is shown via notifications.
• Single Worker Limitation: Remember to run the server with only one worker (--workers 1 if using Uvicorn directly) due to the use of shared in-memory state.

---

🙌 Contributing 🙌

Contributions are welcome! If you have suggestions, bug reports, or want to add features:

1. Fork the repository.
2. Create a new branch (git checkout -b feature/your-feature-name).
3. Make your changes.
4. Commit your changes (git commit -am 'Add some feature').
5. Push to the branch (git push origin feature/your-feature-name).
6. Create a new Pull Request.

Please report any issues using the GitHub Issues tracker.

---

🙏 Acknowledgements 🙏

This project heavily relies on the amazing work by the creators of:

• Real-ESRGAN: https://github.com/xinntao/Real-ESRGAN
• FastAPI: https://fastapi.tiangolo.com/
• BasicSR: https://github.com/xinntao/BasicSR

Give them a star! ⭐