Blurry - Face Privacy Protection - Hackathon Submission

Inspiration

The inspiration for Blurry came from observing how Gen Z users interact with social media. Research shows that 21.7% of young users prefer hiding their faces in profile pictures—not primarily for privacy concerns, but because they feel embarrassed about "trying too hard" or looking too posed. We realized there was a gap between wanting to share memories and feeling comfortable doing so. Traditional face-blurring apps felt too clinical and focused on security rather than self-expression.

With Blurry, we wanted to make privacy protection feel natural and even stylish—transforming the act of hiding faces into a form of creative expression.

What it does

Blurry is a simple yet powerful iOS app that automatically detects faces in photos and applies privacy-protecting effects with just one tap. Users can choose from:

  • Blur effects: Natural, soft blurring that keeps the photo aesthetic
  • Mosaic patterns: Classic pixelation for strong privacy
  • Emoji overlays: Fun, expressive emojis that hide faces while adding personality
  • Solid color fills: Clean, minimalist coverage

All processing happens 100% locally on the device, ensuring that no photo or personal data ever leaves the user’s phone. The app is designed for the modern social media workflow—quick, intuitive, and privacy-first.

How we built it

We built Blurry using modern iOS technologies to ensure both performance and privacy:

Core Technologies

  • SwiftUI: Main interface and user experience
  • MediaPipe Tasks Vision: High-accuracy, real-time face detection
  • Core ML & Core Image: Optimized on-device processing and visual effects
  • RevenueCat & Google Mobile Ads: Subscription and monetization integration
  • iOS Localization: Full multilingual support

Architecture

  • SwiftUI-based simple architecture (instead of MVVM)
  • Singleton patterns where appropriate in actual implementation
  • Local storage only—no cloud services or external APIs
  • Privacy-by-design principles from day one

Development Process

  • Started with user research on social media privacy behaviors
  • Iteratively tested detection accuracy across diverse demographics
  • Migrated from React Native + Expo to native SwiftUI for performance and accuracy
  • Optimized

Built With

Share this project:

Updates