Real-time hand tracking
MediaPipe detects up to two hands and maps 21 landmarks per hand to an interaction model.
Control digital energy with your hands.
NeuroFX Studio is an AI-powered gesture interaction platform that transforms real-time hand tracking into immersive visual effects directly in the browser.
60 FPS
Browser-native loop
21 pts
Landmarks per hand
Local
No camera uploads
Live Studio Preview
2 hands detected
Gesture
Two Hands
Effect
Energy Orb
Confidence
94%
AI running
FPS
62
Stable render loop
Latency
12 ms
GPU delegated model
Events/min
24
Gesture throughput
Camera input is processed locally in your browser.
Real-time FX engine ready for capture.
Product Surface
Built as an interaction platform, not a disposable demo.
Gesture-reactive FX
Pinch, open hand, fist, swipe, push and two-hand energy gestures drive canvas effects.
Performance telemetry
FPS, latency, confidence, session time and interaction volume are visible as product metrics.
AI-ready architecture
A local narrator layer is prepared for future OpenAI integration without exposing API keys.
Privacy by design
No image, video, biometric data or camera stream is uploaded to any server.
Portfolio-grade product
Landing, Studio, dashboard, about page, documentation and Vercel-ready build are included.
Technical Stack
Browser-native computer vision with production web ergonomics.
The camera, model loading, canvas rendering and browser APIs stay inside client components. The static product shell remains compatible with server rendering and Vercel.
Next.js App Router
TypeScript
Tailwind CSS
React
MediaPipe Tasks Vision
Canvas 2D
Framer Motion
Recharts
Vitest
Vercel
Use Cases
Designed for screen recordings, technical walkthroughs and live portfolio reviews.
01
Portfolio demos for computer vision and HCI
02
Spatial-computing experiments in the browser
03
Interactive product launches and technical storytelling
04
Gesture-controlled dashboards and creative tools
05
Education for MediaPipe, Canvas and real-time graphics
Privacy-first camera processing.
This project processes camera input locally in the browser. No image, video, biometric data or camera stream is uploaded to any server.
Este projeto processa a câmera localmente no navegador. Nenhuma imagem, vídeo, dado biométrico ou transmissão da câmera é enviada para servidores.