HeardAI - Enhanced iOS Voice Assistant

🎤 Advanced Voice Assistant with Robust Wake Word Detection & Google Speech Integration

HeardAI is an enhanced iOS voice assistant that provides superior wake word detection, high-accuracy transcription using Google Cloud Speech-to-Text, and seamless Siri integration.

✨ Features

🎯 Enhanced Wake Word Detection

• Reliable Detection: "Hey HeardAI" wake word with audio level monitoring
• Real-time Processing: Continuous audio monitoring with visual feedback
• Noise Handling: Adaptive thresholds for different environments

🎵 Google Speech-to-Text Integration

• High Accuracy: Google Cloud Speech-to-Text API for superior transcription
• Real-time Processing: Live audio capture and conversion
• WAV Format Support: Proper audio format conversion for API compatibility

🤖 Siri Command Execution

• 8 Command Types: Reminders, messages, calls, weather, music, timers, alarms, general
• Smart Parsing: Automatic command type detection and parameter extraction
• Proper Integration: Uses SiriKit instead of problematic URL schemes

📱 Command Confirmation System

• Edit Commands: Review and modify transcribed commands before execution
• User Control: Accept or reject commands with visual feedback
• Command History: Track previous commands for reference

🚀 Quick Start

Prerequisites

• Xcode 14.0 or later
• iOS 16.0 or later
• Physical iOS device (audio features don't work in simulator)
• Google Cloud Speech-to-Text API key

Installation

1. Clone the repository

   git clone https://github.com/Codingbysid/HeardAI1.git
   cd HeardAI1

2. Open in Xcode

   open HeardAI.xcodeproj

3. Configure API Key

   • The Google Speech API key is already configured in Info.plist
   • Current key: 0ac4a773d6788813e45539a3ee8c4bdf6767faa8

4. Build and Run

   • Select your physical device
   • Build and run the project (⌘+R)

📱 Usage

1. Launch the app and grant microphone permissions
2. Say "Hey HeardAI" to activate the assistant
3. Speak your command clearly after the wake word
4. Review and edit the transcribed command if needed
5. Accept or reject the command for execution

🎯 Supported Commands

Command Type	Examples
Reminders	"remind me to call mom", "set reminder for meeting"
Messages	"send message to John", "text Sarah hello"
Calls	"call dad", "phone mom"
Weather	"what's the weather", "weather forecast"
Music	"play music", "play Bohemian Rhapsody by Queen"
Timers	"set timer for 5 minutes", "timer 10 minutes"
Alarms	"set alarm for 7 AM", "wake me up at 8"
General	"what time is it", "open settings"

🛠️ Technical Architecture

Core Services

• AudioManager: Handles wake word detection and audio processing
• WhisperService: Google Speech-to-Text integration
• SiriService: Command parsing and Siri execution
• SpeechRecognizer: On-device speech recognition fallback

Key Technologies

• SwiftUI: Modern declarative UI framework
• AVFoundation: Audio capture and processing
• Google Cloud Speech-to-Text: High-accuracy transcription
• SiriKit: Proper Siri integration with intents
• Combine: Reactive programming for state management

🔧 Configuration

Required Permissions

<key>NSMicrophoneUsageDescription</key>
<string>HeardAI needs microphone access to listen for voice commands and the wake word "Hey HeardAI".</string>

<key>NSSpeechRecognitionUsageDescription</key>
<string>HeardAI uses speech recognition to accurately transcribe your voice commands before sending them to Siri.</string>

Build Settings

• iOS Deployment Target: 16.0
• Swift Language Version: Swift 5
• Enable Bitcode: No

🧪 Testing

Manual Testing

1. Wake word detection in quiet environment
2. Wake word detection in noisy environment
3. Command transcription accuracy
4. Siri command execution
5. Command confirmation flow

Demo Mode

• Use the "Test Command (Demo)" button to simulate wake word detection
• Test the command confirmation system without voice input

🚨 Troubleshooting

Common Issues

"Speech recognition not available"

• Ensure testing on physical device (not simulator)
• Check speech recognition permissions
• Verify internet connectivity

"Google Speech API error"

• Verify API key is valid and active
• Check network connectivity
• Ensure API quotas haven't been exceeded

"Wake word not detected"

• Speak clearly and at normal volume
• Ensure microphone permissions are granted
• Check that the app is in the foreground

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Google Cloud for the Speech-to-Text API
• Apple for Speech Framework and SiriKit
• SwiftUI community for UI best practices

📞 Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions

---

Made with ❤️ for the iOS community

⭐ Star this repository if you find it helpful!