Desktop application to identify bird vocalisations in lengthy audio files. Uses either BirdNET or a native AI model tuned for the calls of nocturnal migrants.
Author: Matthew Kirkland
- Uses two Machine Learning models to identify audio files based on the user's needs: BirdNET and the Nocmig model
- Supports audio input files such as WAV, MP3, MP4/M4A, AAC, Opus, Ogg, and FLAC
- Audio analysis can run in the background while exploring the application
- Tailor species detection based on the season, time of day, or a custom list of species
- Program can reduce background noise to make avian sounds more audible
- ...and more
Visit https://chirpity.mattkirkland.co.uk for platform specific installation instructions - Chirpity binaries are available for both Windows and Mac platforms. Linux users will need to run the application from source, as described below.
First, clone the project and install all dependencies:
git clone https://github.com/Mattk70/Chirpity-Electron
cd Chirpity-Electron
Chirpity depends on Node.js, follow the link for the download and installation instructions. Once installed, run:
npm install
Next, launch the app with:
npm start
Initialize the source directory with:
npm init
Now, install project dependencies with:
npm install --save-dev
After that, build a windows msi installer with:
npm run export
The resulting application will be saved in the "dist" folder.