Sign Me Up!: Learn ASL with AI

Project Overview

Sign Me Up is a privacy-focused ASL learning platform that uses MediaPipe Hands and vanilla JavaScript to track your hand gestures directly through your webcam. It provides real-time feedback without requiring any accounts, logins, or servers. All progress is saved locally on your browser using localStorage, so nothing ever leaves your device.

Homepage Screenshot

Introductory Modules

Users begin with basic signs like greetings, yes/no questions, and everyday phrases. Each module tracks your gesture progress locally and requires successful repetition to move forward. Tutorials for each sign are coming soon and will include visual walkthroughs to help new learners.

Intro Module Lessons

Modular Lesson Cards

Lessons are organized into clean, mobile-friendly cards that guide you through specific signing categories. Each card brings you to an interactive camera session with instructions based on the context of that sign. The design emphasizes clarity, even on smaller devices.

Lesson Cards UI

Camera-Based Detection

When you begin a lesson, the app accesses your webcam and uses MediaPipe to track hand movement live. The setup is intentionally lightweight and fully browser-based β€” no downloads, no backend, and no uploads β€” to make it accessible across devices.

Camera Starting Screen

Real-Time Feedback

As you perform each sign, the app visually confirms whether the gesture is correct with a βœ… directly on screen. This feedback loop helps users adjust quickly and build muscle memory with minimal delay or confusion.

Positive Detection Feedback

When Detection Fails

If the system doesn’t recognize a sign, a ❌ appears on the screen. Users can immediately retry without penalty, and logic has been added to prevent accidental triggers. This makes the experience more forgiving, especially during early attempts or noisy environments.

Negative Detection Feedback