pyvoice 0.1.0

A real-time speech-to-text transcription tool using machine learning (NumPy), PyQt6, and faster-whisper.

PyVoice

PyVoice is a real-time speech-to-text transcription tool using machine learning. It provides a graphical user interface for recording audio and transcribing it into text.

Features

• Real-time audio recording
• Live audio waveform visualization
• Speech-to-text transcription using the Whisper model
• User-friendly GUI built with PyQt6

Requirements

The script automatically checks for and installs the following required packages:

• numpy
• sounddevice
• librosa
• faster_whisper
• PyQt6
• pyqtgraph

Installation

1. Ensure you have Python installed on your system.
2. Download the PyVoice script file.
3. Run the script. It will automatically check for and install required packages.

Usage

1. Run the PyVoice script to launch the application.
2. Click "Start Recording" to begin audio capture.
3. Speak into your microphone.
4. Click "Stop Recording" when you're done speaking.
5. PyVoice will process the audio and display the transcription.

Key Components

• AudioRecorder: Handles real-time audio recording.
• TranscriptionWorker: Processes recorded audio and generates transcriptions.
• MainWindow: Manages the graphical user interface and coordinates recording and transcription processes.

Notes

• PyVoice uses the "small" Whisper model for transcription.
• Audio is processed at a 16kHz sample rate.
• The waveform display updates in real-time during recording.