frame-processor

React Native Vision Camera v5 (Nitro) plugin for Google ML Kit — text recognition (OCR) and barcode scanning today, each independently installable, with more ML Kit features on the roadmap.

Real-time body pose detection for React Native & Expo using Google MediaPipe. 33 body landmarks, VisionCamera frame processors, GPU acceleration, photo detection and an Expo config plugin. New Architecture (TurboModules).

📸 The Vision Camera equivalent for Flutter. 🚀 High-performance, zero-overhead native bindings using the Nitro SDK. Real-time GPU-accelerated video pipeline, custom GLSL shaders, and ultra-low latency FFI for iOS (AVFoundation) and Android (Camera2).

Face Detector POC is a React Native demo app using Vision Camera with Worklets Core to perform real-time, on-device face detection. It leverages Google ML Kit and iOS Vision API to detect multiple faces, landmarks, and contours efficiently on both iOS and Android.

Android app that turns the camera into a language-learning surface: on-device object detection with MediaPipe + VisionCamera 5 and a Kotlin Nitro module. No backend, no frame leaves the phone.