60f18f8e71
- Implement SpeechToText class with Docker-based transcription capabilities - Add wake word detection using OpenWakeWord and fast-whisper models - Create Dockerfile for speech processing container - Develop comprehensive test suite for speech recognition functionality - Include audio processing and event-driven transcription features
39 lines
1.1 KiB
Docker
39 lines
1.1 KiB
Docker
FROM python:3.10-slim
|
|
|
|
# Install system dependencies
|
|
RUN apt-get update && apt-get install -y \
|
|
git \
|
|
build-essential \
|
|
portaudio19-dev \
|
|
python3-pyaudio \
|
|
&& rm -rf /var/lib/apt/lists/*
|
|
|
|
# Install fast-whisper and its dependencies
|
|
RUN pip install --no-cache-dir torch torchaudio --index-url https://download.pytorch.org/whl/cpu
|
|
RUN pip install --no-cache-dir fast-whisper
|
|
|
|
# Install wake word detection
|
|
RUN pip install --no-cache-dir openwakeword pyaudio sounddevice
|
|
|
|
# Create directories
|
|
RUN mkdir -p /models /audio
|
|
|
|
# Download the base model by default
|
|
RUN python -c "from faster_whisper import WhisperModel; WhisperModel.download_model('base.en', cache_dir='/models')"
|
|
|
|
# Download OpenWakeWord models
|
|
RUN mkdir -p /models/wake_word && \
|
|
python -c "import openwakeword; openwakeword.download_models(['hey_jarvis', 'ok_google', 'alexa'], '/models/wake_word')"
|
|
|
|
WORKDIR /app
|
|
|
|
# Copy the wake word detection script
|
|
COPY wake_word_detector.py .
|
|
|
|
# Set environment variables
|
|
ENV WHISPER_MODEL_PATH=/models
|
|
ENV WAKEWORD_MODEL_PATH=/models/wake_word
|
|
ENV PYTHONUNBUFFERED=1
|
|
|
|
# Run the wake word detection service
|
|
CMD ["python", "wake_word_detector.py"] |