Perception Driver Audio

fp_perception_driver_audio provides audio input and output drivers for ROS2 using the PortAudio library. It defines two main drivers built on a shared DriverBase:

• MicrophoneAudioDriver: Captures audio from the system microphone
• SpeakerAudioDriver: Plays back audio to the system's default speaker

These drivers are intended to be used within the fp_perception ROS2 package and integrate with the event logging and configuration system defined by the DriverBase interface.

Features

• Capture audio in real-time from the microphone
• Play audio buffers through the default output device
• Built on PortAudio for cross-platform low-latency audio I/O
• Uses typed driver interfaces loaded through pluginlib
• Publishes standard ROS 2 diagnostics on /diagnostics when use_diagnostics=true
• Easily extendable for streaming, audio topics, or speech modules

Dependencies

Install PortAudio on Ubuntu:

sudo apt update
sudo apt install libportaudio2 libportaudio-dev python3-pyaudio

To use find_device.py to identify availabe devices, install

Parameters

driver:
  audio:
    MicrophoneAudioDriver:
      name: MicrophoneAudioDriver
      device_id: 0
    SpeakerAudioDriver:
      name: SpeakerAudioDriver
      device_id: 0

Usage

This module is not meant to be run as a standalone node. Instead, it is designed to be loaded as a plugin or instantiated via your own driver management system in ROS2.

Example integration:

auto driver = std::make_shared<fp_perception::MicrophoneAudioDriver>();
driver->initialize(node);
auto data = driver->readChunk();

To play audio using SpeakerAudioDriver:

fp_perception::SpeakerAudioDriver speaker;
speaker.initialize(node);
speaker.play(audio_data);

Notes

• Microphone capture and speaker playback use PortAudio callbacks.
• The perception server can route diagnostics to /diagnostics through diagnostic_updater.

To-Do

• [ ] Add long-running runtime validation for simultaneous capture and playback.