Wake Me

AI-based browser component library that detects clapping sounds or finger snaps. Using a TensorFlow.js-based machine learning model, it accurately analyzes sounds in real-time.

This project is provided by LLAMI Team

Read this in other languages: English, 日本語

Online Demo

Try out Wake Me in your browser: Online Demo

⚠️ Audio Setup Recommendation:

• For best results, please use external speakers and microphone
• The detection may not work properly with earphones/earbuds
• Works on both desktop and mobile devices!

Use Cases

• 🎙️ Voice Recognition AI Systems: As a wake word alternative to activate AI using claps or snaps
• 🎮 Game Control: Sound-based interaction for hands-free gaming scenarios
• 🎭 Interactive Presentations: Trigger slide transitions or animations during presentations
• 🖥️ Video Conference Assistant: Signal for speaking turns or attention
• 🎨 Digital Art Installations: Interface for audience-artwork interaction
• 👥 Accessibility Enhancement: Alternative input method for users with physical limitations
• 🤖 Smart Home Control: Simple trigger method for IoT device control
• 📱 Mobile App Control: App control in hands-busy situations

Key Features

• 🤖 High-performance AI model powered by TensorFlow.js
• ⚡ Real-time sound detection and analysis
• 🎯 High-accuracy clap/snap recognition
• 🪶 Lightweight and easy to use

Framework Support

WakeMe is designed to be framework-agnostic! While it's primarily a React component, you can use it with:

• React
• Vanilla JavaScript
• Angular
• Vue.js
• Any other JavaScript framework

React Usage

import { WakeMe } from "wake-me";

function App() {
  return <WakeMe onSnap={() => console.log("Snap detected!")} />;
}

Vanilla JavaScript Usage

<script src="https://cdn.jsdelivr.net/npm/wake-me@latest/dist/vanilla/vanilla.global.js"></script>
<script>
  const wakeMe = new WakeMe({
    onSnap: () => {
      // Handle snap event
      console.log("Snap!");
    },
    onNoise: (score) => {
      // Handle noise level
      console.log("Noise level:", score);
    },
  });

  // Start listening
  wakeMe.init();

  // Clean up when done
  // wakeMe.destroy();
</script>

Installation

npm install wake-me
# or
yarn add wake-me

Usage

import { WakeMe } from "wake-me";

function App() {
  return <WakeMe onSnap={() => console.log("Snap detected!")} />;
}

Props

| Prop | Type | Required | Description | | ----------------------- | ----------------------- | -------- | ----------------------------------------------------------------------------------------------------------- | | onSnap | () => void | Optional | Callback when snap is detected | | onNoise | (score: number) => void | Optional | Callback when noise is detected | | modelBaseUrl | string | Optional | Base URL for model files (default: '/snap/') | | tfScriptUrl | string | Optional | TensorFlow.js script URL | | speechCommandsScriptUrl | string | Optional | Speech Commands script URL | | snapThreshold | number | Optional | Detection sensitivity threshold (default: 0.95). Can be lowered to 0.9 if detection is not sensitive enough |

Model File Setup

The following files must be present in the modelBaseUrl for the component to work properly:

• model.json
• metadata.json
• *.bin files

Embedding Scripts and Model Files (Next.js Example)

You can embed the required scripts and model files directly in your project. Here's an example using Next.js:

1. Download Required Files

Download the following files:

• TensorFlow.js: https://cdn.jsdelivr.net/npm/@tensorflow/tfjs/dist/tf.min.js
• Speech Commands: https://cdn.jsdelivr.net/npm/@tensorflow-models/speech-commands/dist/speech-commands.min.js
• Model files:
  • model.json
  • metadata.json
  • *.bin files

2. File Placement

In a Next.js project, place static files in the public directory:

your-nextjs-project/
├── public/
│   ├── scripts/
│   │   ├── tf.min.js
│   │   └── speech-commands.min.js
│   └── models/
│       ├── model.json
│       ├── metadata.json
│       └── *.bin files

3. Using in Component

import { WakeMe } from "wake-me";

function App() {
  return (
    <WakeMe
      onSnap={() => console.log("Snap detected!")}
      // Set script URLs
      tfScriptUrl="/scripts/tf.min.js"
      speechCommandsScriptUrl="/scripts/speech-commands.min.js"
      // Set model files path
      modelBaseUrl="/models/"
    />
  );
}

In Next.js, files in the public directory are accessible from the root URL ('/').

License

MIT