Amin-moniry-pr7 / Speech-to-Text-Transcription

This project automates audio processing by removing silence, transcribing speech to text, and storing the output in an SQLite database. It supports multiple audio formats and leverages Google Speech Recognition for high accuracy.

🎤 Audio Processing & Transcription Tool

A powerful Python-based audio processing tool that automatically removes silence from audio files and transcribes speech to text with high accuracy.

🌟 Features

• Intelligent Silence Removal: Automatically detects and removes silence segments from audio files
• Multi-format Support: Works with WAV, MP3, M4A, OGG, and FLAC audio formats
• Speech-to-Text Transcription: Uses Google Speech Recognition for accurate transcription
• Database Integration: SQLite database stores processing metadata and transcriptions
• Duration Tracking: Monitors original vs. processed audio durations
• Configurable Parameters: Customizable silence detection thresholds
• Robust Error Handling: Comprehensive error management and validation

Quick Start

Prerequisites

• Python 3.7 or higher
• Internet connection (for Google Speech Recognition)

Installation

1. Clone the repository

   git clone https://github.com/Amin-moniry-pr7/Audio-Processing-Transcription.git
   cd Audio-Processing-Transcription

2. Install dependencies

   pip install -r requirements.txt

3. Run the application

   python CONVERT_AUDIO_TO_TEXT_AND_REMOVE_SILENCE.py

📋 Usage

When you run the main script, you'll be prompted to provide:

1. Audio file path: Path to your input audio file
2. Output file path: Where to save the transcription (e.g., output.txt)
3. Language code: Language for transcription (e.g., en-US, de-DE, fr-FR)
4. Silence parameters:
   • Minimum silence length (milliseconds, e.g., 1000)
   • Silence threshold (dB, e.g., -40.0)

Example Usage

Enter the path to an audio file: ./audio/interview.mp3
Enter the path to save the transcription: ./transcriptions/interview.txt
Enter the language code: en-US
Minimum silence length in milliseconds: 1000
Silence threshold in dB: -40.0

📁 Project Structure

📁 Audio-Processing-Transcription/
├── 📜 CONVERT_AUDIO_TO_TEXT_AND_REMOVE_SILENCE.py  # Main entry point
├── 📜 Database_And_prepare_audio.py                # Database operations & audio preparation
├── 📜 Remove_silence_and_mesuere.py               # Silence removal & duration measurement
├── 📜 Speech_and_transcribe.py                    # Speech-to-text processing
├── 📜 requirements.txt                            # Project dependencies
├── 📜 LICENSE                                     # Apache 2.0 License
├── 📜 .gitignore                                  # Git ignore rules
└── 📜 README.md                                   # This file

🗄️ Database Schema

The application creates an SQLite database (PODCAST.db) with the following structure:

Column	Type	Description
id	INTEGER	Primary key (auto-increment)
input_path	TEXT	Original audio file path
output_path	TEXT	Transcription output file path
language	TEXT	Language code used for transcription
original_duration	REAL	Duration of original audio (seconds)
processed_duration	REAL	Duration after silence removal
transcription	TEXT	Full transcription text
created_at	TIMESTAMP	Processing timestamp

🔧 Configuration

Supported Audio Formats

• WAV (recommended for best quality)
• MP3 (most common format)
• M4A (Apple format)
• OGG (open-source format)
• FLAC (lossless compression)

Language Codes

Common language codes for transcription:

• en-US - English (US)
• en-GB - English (UK)
• de-DE - German
• fr-FR - French
• es-ES - Spanish
• it-IT - Italian
• ja-JP - Japanese
• ko-KR - Korean

Silence Detection Parameters

• Minimum Silence Length: Minimum duration (in ms) to consider as silence
  • Typical range: 500-2000ms
  • Lower values = more aggressive silence removal
• Silence Threshold: Volume level (in dB) below which audio is considered silence
  • Typical range: -30 to -50 dB
  • Lower values = more sensitive silence detection

📊 Output Files

The tool generates several files during processing:

1. Processed Audio: [original_name]_no_silence.wav - Audio with silence removed
2. Transcription File: User-specified text file containing the full transcription
3. Database Record: Entry in PODCAST.db with all processing metadata

🛠️ Dependencies

pydub>=0.25.1
SpeechRecognition>=3.8.1

Additional system requirements:

• FFmpeg (for audio format conversion)
• Internet connection (for Google Speech Recognition API)

📈 Performance Tips

• Use WAV format for fastest processing (no conversion needed)
• Optimize silence parameters based on your audio content:
  • Podcasts/interviews: 1000ms silence, -40dB threshold
  • Music: 500ms silence, -50dB threshold
  • Noisy environments: 2000ms silence, -30dB threshold

🔍 Troubleshooting

Common Issues

"Error: File not found"

• Verify the audio file path is correct
• Ensure the file exists and is readable

"Invalid input" for silence parameters

• Ensure minimum silence length is a positive integer
• Ensure silence threshold is a negative number

Transcription errors

• Check your internet connection
• Verify the language code is correct
• Ensure the audio quality is sufficient for recognition

Audio format not supported

• Install FFmpeg for additional format support
• Convert audio to WAV format manually if needed

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request. For major changes, please open an issue first to discuss what you would like to change.

📄 License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

👨‍💻 Author

Amin Moniry

• GitHub: @Amin-moniry-pr7

🙏 Acknowledgments

• Google Speech Recognition API for transcription services
• PyDub library for audio processing
• SQLite for lightweight database functionality

---

⭐ If you found this project helpful, please consider giving it a star! ⭐