Gesto AI

Real-time PSL → Urdu Speech

Gesto AI

Turning Pakistan Sign Language into spoken Urdu — instantly — so deaf and mute voices are finally heard.

Try Live Demo Why it matters

Live Demo

Sign it. We say it.

Start the camera, sign in front of it, and watch Gesto AI recognize your gesture in real time.

Camera is off

Tap “Start Camera” to begin

Server disconnected

Recognized Gesture

—

Waiting for camera…

Confidence 0%

Recent Gestures

  • Recognized gestures will appear here.

The Problem

9 million Pakistanis sign. Almost nobody hears them.

Pakistan has one of the largest deaf populations in the world, yet sign-language interpreters are scarce and most hearing people cannot understand PSL. Everyday moments — a doctor's visit, a market purchase, an emergency — become walls of silence.

Gesto AI bridges that gap: a camera watches the signer's hands, an AI model reads the gesture, and the app speaks it aloud in Urdu — in real time, on ordinary devices, with no special hardware.

  1. Camera

    Captures the signer's hand gestures through any phone or laptop webcam.

  2. AI Model

    MediaPipe tracks the hand while a TensorFlow classifier names the PSL gesture.

  3. Urdu Speech

    The recognized gesture is spoken aloud in Urdu, closing the conversation gap.

The Team

The minds behind the magic.

Muhammad Ghuzaif Arif

Muhammad Ghuzaif Arif

ML Engineer

Trains and tunes the PSL gesture classifier on top of MediaPipe hand landmarks.

Adeel Shahbaz

Adeel Shahbaz

Backend Engineer

Builds the FastAPI inference server and the real-time WebSocket prediction pipeline.

Ishmal Nayyab

Ishmal Nayyab

Frontend & UX

Crafts the accessible, judge-friendly interface and the Urdu voice experience.

Tech Stack

Small stack, real-time results.