Upscaile

📌 Motivation

Many AI-powered image enhancement sites limit users with credit systems or paid subscriptions, making it difficult to freely upscale images. I built Upscaile to solve this — a free, intuitive platform that uses AI to upscale and enhance your images up to 4× their original resolution.

What started as a simple upload form and button gradually evolved into a full-featured web app. I experimented with frontend frameworks, APIs, and design iterations to optimize performance and usability. Today, Upscaile integrates the Hugging Face finegrain-image-enhancer API for enhancement, giving users a seamless and accessible experience.

tl;dr: Tired of paywalls and clunky tools for AI image upscaling? Upscaile lets you freely enhance images 4× their resolution — fast, clean, and free.

⚙️ Tech Stack

• Frontend: HTML, CSS, JavaScript, Bootstrap
• Backend: Flask (Python), Redis (upscale count)
• APIs/Services:
  • Hugging Face Spaces — AI upscaling (finegrain-image-enhancer)
  • Cloudinary — Image upload/storage
  • SightEngine — Image moderation
  • Render — Web app deployment

🔄 How It Works

1. Upload Image: User uploads an image via the web interface (drag & drop or file select).
2. Choose Settings: Select the desired upscale factor (2x, 3x, or 4x) and submit.
3. Image Moderation: The image is checked for safety using SightEngine.
4. Cloudinary Upload: The image is uploaded to Cloudinary to generate a public URL.
5. AI Enhancement: The Hugging Face API processes the image using advanced AI upscaling models.
6. Result Display: The enhanced image is shown side-by-side with the original, with a download option. The original upload is deleted from Cloudinary for privacy.

All of this happens in just a few seconds.

🚀 How to Use

1. Upload your image (drag & drop or file select).
2. Choose your upscale factor (2x, 3x, or 4x).
3. Click Submit.
4. Wait a few seconds — your upscaled image will be displayed with a download link.

✨ Features

• Modern, section-based UI with gradients and clear step-by-step flow
• Side-by-side preview of original and enhanced images
• Progress bar feedback during upscaling
• Product demo carousel and "How It Works" explainer section
• No sign-up, no paywall, no watermark

🛠 Development Process & Updates

• Modularization: Refactored backend into service modules (Cloudinary, Hugging Face, moderation, Redis, file utils) for maintainability.
• API Migration: Utilized Hugging Face finegrain-image-enhancer API for reliable, high-quality upscaling.
• Frontend Overhaul: Moved CSS to a static file, improved layout, added gradients, and made the UI responsive and visually appealing.
• User Experience: Added progress bar, instant preview of uploaded image, and clear feedback for errors and success.
• Security & Privacy: All images are moderated and deleted from Cloudinary after processing.

🛡 License

This project follows the MIT License, as specified in the Hugging Face Space.

Try it out yourself at https://upscaile.onrender.com!

🏃‍♂️ Local Development

1. Clone the repo and install dependencies:

   git clone https://github.com/AdiBak/upscaile.git
   cd upscaile
   pip install -r requirements.txt

2. Set up your environment variables in .github/envvars.env (see setup.py for details).
3. Run the app locally:

   python main.py

4. Visit http://localhost:5000 in your browser.