Blog

BirdNET-Pi: Using AI to identify Birds from their Songs.

Wouldn’t it be fantastic to be able to identify birds from their song? Well it turns out that you can.

BirdWeather
BirdWeather is a great citizen scientist initiative that uses volunteer stations to continuously listen for bird songs all over the world. The stations use machine learning to analyse these sounds and identify birds based on their song. BirdWeather is a living library of bird vocalizations. It also provides an indication of bird visit frequency and populations around a station, apart from those which keep quiet! You can play back the recorded songs and bird detections are shown in real time on the web site.

The artificial neural network used to analyse the recorded sound and extract and identify bird songs is called BirdNET.

BirdNET
BirdNET is a joint research project between The Cornell Lab of Ornithology and the Chemnitz University of Technology that is “focused on the detection and classification of avian sounds using machine learning … [with the aim to] assist experts and citizen scientist in their work of monitoring and protecting our birds”.

Detection Stations
There are currently only 12 stations in Australia (including me), all based on the Raspberry Pi single board computer. You can learn how to configure your own station in our latest article on Medium – BirdNET-Pi: Using AI to identify Birds from their Songs.

There are a number of libraries that make working with the LSM9DS1 IMU (Inertial Measurement Unit) easy. In Part 5 of our series on writing your own flight controller software we will focus on the IMU and in particular explain the best ways to convert the gyro rate and accelerometer force data to a roll and pitch angle. The yaw angle is then calculated using the pitch and roll and magnetometer data. This discussion, is in the context of building a drone flight controller based on the Arduino Nano 33 BLE.

Our follow up part in this series will explain how to calibrate the LSM9DS1 IMU and store the results plus use sensor fusion to improve the accuracy of the angles and overcome sensor shortfalls.

Read the complete article on Medium.