In addition to the BirdNET-Pi, we also have other tools, such as the BirdNET-Analyzer.
If we have enough time, we will work with real-world bird sound recordings collected in advance using AudioMoth devices. The AudioMoth is a compact, low-power acoustic recorder that is widely used in ecological and bioacoustic research to capture environmental sounds over extended periods of time.
These recordings will be analyzed using the BirdNET-Analyzer, a tool developed to automatically detect and classify bird species from audio data.
Steps:
-
Download and set up the BirdNET-Analyzer from the GitHub repository.
-
Upload the pre-recorded audio files from the AudioMoth devices to the BirdNET-Analyzer.
- Note: We will only analyze some parts of the audio files, as analyzing the entire recordings would be too time-consuming.
- If the file is too long, deselect the option
Generate spectrogram.
-
Run automated species detection.
-
Explore the results, including timestamps, confidence scores, and predicted species.
This hands-on session provides practical insight into how bioacoustic data is processed and analyzed using AI, and demonstrates how such tools can support research, monitoring, and conservation efforts.