Empowering Communication Without Limits
SignBridge AI

American Sign Language โ†’ Text โ†’ Speech ยท Multilingual Support
MobileNetV2 ยท MediaPipe ยท WebSocket ยท Browser Camera ยท Cloud Ready โ˜๏ธ

MobileNetV2
MediaPipe Hands
Zero Lag WebSocket
Browser Camera API
CAMERA_FEED ยท Browser โ†’ Server โ†’ Canvas ยท MobileNetV2 + MediaPipe

Click below to start your camera.
Your browser will ask for camera permission.

FLAT HAND space
PINCH end+speak
THUMBS UP autocorrect
Qโ€“T word suggest
1โ€“5 pick correction
BKSP/Z delete
C clear all
System Information

How Sign-Bridge AI Works

Powered by MobileNetV2 and MediaPipe Hands for real-time ASL recognition

๐Ÿง 

Model Architecture

  • Base: MobileNetV2 (Fine-Tuned layers 100โ€“155)
  • Input Shape: 128 ร— 128 ร— 3
  • Output Classes: Aโ€“Z + SPACE + DOT + AUTOCORRECT
  • Feature Vector: n=1280 (flattened)
  • Classifier: Fully Connected โ†’ Softmax
โš™๏ธ

Recognition Settings

  • Confidence Threshold: 0.75
  • Smoothing Window: 10 frames
  • Letter Hold Duration: 1.0 sec
  • Special Gesture Hold: 1.2 sec
  • WebSocket Frame Rate: adaptive
๐Ÿ“ก

Live System Status

WebSocket Connecting โ€ฆ
ML Model โ€”
Autocorrect โ€”
TTS Engine gTTS Cloud โœ“
๐Ÿงฉ

Tech Stack

Python 3.10
TensorFlow / Keras
MobileNetV2
MediaPipe Hands
OpenCV Headless
FastAPI + Uvicorn
WebSocket
gTTS Cloud TTS
LanguageTool NLP
Browser Camera API
ASL Alphabet Chart
ASL Sign Language Alphabet
MobileNetV2 Architecture
Model Architecture Diagram
โœ‹

Special Gesture Commands

  • Flat Hand (1.2s hold) โ†’ Insert SPACE
  • Pinch gesture โ†’ End sentence + Speak
  • Thumbs Up โ†’ Trigger Autocorrect
  • Qโ€“T keys โ†’ Select word suggestion
  • 1โ€“5 keys โ†’ Pick autocorrect variant
  • Z / Backspace โ†’ Delete last letter
  • C key โ†’ Clear all text