Lateral Thinking Puzzle Game with Voice Interaction

Overview

This project is an interactive lateral thinking puzzle game that combines voice interaction, text-to-speech, and AI-powered responses. Players can solve puzzles through both text and voice input, with the system providing voice feedback using ElevenLabs' voice cloning technology.

Features

• Interactive terminal-based puzzle interface
• Voice input support for puzzle solving
• Text-to-speech responses using ElevenLabs
• Voice cloning capability after multiple interactions
• Dynamic AI responses using OpenAI
• Real-time audio processing and transcription

Technical Stack

• Frontend: jQuery Terminal
• Backend: Express.js
• APIs:
  • ElevenLabs for text-to-speech and voice cloning
  • OpenAI for transcription and AI responses
• Audio Processing: WAV format handling

Key Components

1. Puzzle Engine (Reference: script.js lines 101-114)

   • Manages puzzle flow and user interactions
   • Handles correct answer detection
   • Provides contextual AI responses

2. Voice Processing (Reference: shared/audio.js lines 11-24)

   • Records user voice input
   • Saves audio in WAV format
   • Handles audio playback

3. AI Integration (Reference: shared/openai.js lines 17-25)

   • Transcribes voice input to text
   • Generates contextual responses
   • Manages conversation flow

4. Voice Synthesis (Reference: shared/elevenlab.js lines 21-36)

   • Converts AI responses to speech
   • Handles voice cloning after sufficient samples
   • Manages audio file generation

Setup

1. Install dependencies:

npm install

2. Configure environment variables:

OPENAI_API_KEY=your_key
ELEVEN_LABS_API_KEY=your_key

3. Start the server:

npm run 1

Project Structure

• /public_ver22: Frontend assets and client-side code
• /shared: Server-side utilities and API integrations
• /audio: Dynamic audio file storage

Voice Cloning Feature

The system collects user voice samples during puzzle interactions. After 4 samples (configurable), it creates a personalized voice clone for responses.

Reference: index.js lines 25-29 for voice cloning configuration.

Note

This project requires active API keys and proper audio directory permissions to function correctly. The voice cloning feature needs multiple interactions to gather sufficient voice samples.