Command Line Interface for batdetect2

This repository provides a Python script to enable command line interface (CLI) usage of batdetect2 (https://github.com/macaodha/batdetect2) for batch analysis of large audio datasets in bat monitoring programs and research studies.

It adds some extra functionality to batdetect2 that proved helpful when analyzing large amounts of audio data from an acoustic bat monitoring program in German forests (https://doi.org/10.5281/zenodo.14535158). The development of the CLI approach for batdetect2 was inspired by the great experience with the CLI usage of the BirdNET Analyzer (https://github.com/BirdNET-Team/BirdNET-Analyzer).

The functionality of the script includes:

• Analyzing large amounts of audio data in a controlled and easy way via the command line interface in batches.
• Identifying and skipping audio files that have already been analyzed before (e.g., when restarting your computer after a crash).
• Cutting the audio data into batches per study site instead of using all sites mixed.
• Merging the batdetect2 output per study site instead of storing one single CSV per audio file.
• Adding a row containing "no calls detected" to the batdetect2 output if no call was identified in an audio file.

Install batdetect2

Follow the instructions for installing batdetect2 from https://github.com/macaodha/batdetect2.

Usage of the Command Line Interface

Given that batdetect2 is installed, you can use the batdetect2_cli.py script to start a batch analysis of folders containing ultrasonic audio recordings.

The script is designed to run batdetect2 on an input audio folder (default: ./01_AUDIO_DATA), including one subfolder per study site. The subfolders are named after the ID of the study site (e.g., ./01_AUDIO_DATA/FOREST_01). The script will always use the last folder from the path to identify the site ID (here: "FOREST_01").

1. Open your command line interface (e.g., Anaconda Powershell Prompt).

2. Run the following lines by copying them one by one into the command line interface and hitting "enter".

3. Optional: Activate the batdetect2 Python environment.

   conda activate batdetect2

4. Navigate to the folder where the batdetect2_cli.py script is stored on your computer.

   cd C:\path\to\project\folder

5. Run the batdetect2_cli.py script with arguments for the relative paths to your audio data (--audio-root), the output directory (--output-dir), and the batch size (default = 100).

   python batdetect2_cli.py --audio-root "01_AUDIO_DATA" --output-dir "02_BATDETECT2" --batch-size 100

You will receive one CSV file per study site, named as follows: batdetect2_FOREST_01.csv in the output directory.

The code is free to use and adapt, however, please respect the license of batdetect2!