kurtschwarz/birdy

A realtime distributed acoustic bird classification system using BirdNET.

Birdy

Architecture

@birdy/recorder

The @birdy/recorder service is responsible for collecting audio from a single microphone in a specific location.

• You should deploy as many @birdy/recorder services as you need.
• View the 📕 Recorder Service Docs for details about features, configuration, deployment, etc.

@birdy/collector

The @birdy/collector service is responsible for gathering the raw recordings produced by the @birdy/recorder services and storing them in Minio.

• You should only need to deploy a single @birdy/collector instance.
• View the 📕 Collector Service Docs for details about features, configuration, deployment, etc.

@birdy/analyzer

This is a future machine learning service that will use BirdNET to identify birds via the recordings recorded by @birdy/recorder and collected by @birdy/collector.

• View the 📕 Analyzer Service Docs for details about features, configuration, deployment, etc.

@birdy/api

This is a future api to interact with the Birdy data.

@birdy/app

This is a future web ui/app to view the data collected by Birdy.

Developing

macOS

To develop Birdy on macOS you'll need to install the pulseaudio package to allow us to listen to the microphone inside of Docker.

1. brew install pulseaudio
2. vim /opt/homebrew/Cellar/pulseaudio/17.0/etc/pulse/default.pa
   1. find load-module module-native-protocol-tcp and uncomment it, and then add auth-ip-acl=127.0.0.1;192.168.0.0/24 after it, example:

      ### Network access (may be configured with paprefs, so leave this commented
      ### here if you plan to use paprefs)
      #load-module module-esound-protocol-tcp
      load-module module-native-protocol-tcp auth-ip-acl=127.0.0.1;192.168.0.0/24
      

3. start the pulseaudio server by running pulseaudio --exit-idle-time=-1 --daemon