Mitwelten Audio Uploader

Upload Audio Moth recordings to storage for ML inference tasks.

Build

Use Nuitka to build the application bundles / binaries.

python -m nuitka --standalone --macos-create-app-bundle --enable-plugin=pyside6 \
    --nofollow-import-to=IPython --nofollow-import-to=Pillow  --nofollow-import-to=matplotlib \
    src/uploader_app.py

Usage

• Attach storage medium
• Open Mitwelten Audio Uploader application
• Add a root device (a record pointing to your storage medium). This step will add a file to the target medium containg a UUID, for later identification of that storage medium. You can also add a path in you local filesystem as root device.
• Select a path for indexing: This path has to be a member of on of you root devices. This step will scan the selected path recursively for audio moth recordings and create a records for each recording in a local database (stored in you local user profile)
• To continue processing files you need to sign into your mitwelten account, by clicking sign in.
• Metadata of indexed paths needs to be extracted first.
• Once you start the upload process, files with extracted metadata (ready to upload ) will be uploaded to remote storage (the remote storage location is determined by the back end).

Authors and acknowledgment

mitwelten.org

Project status

This project is actively developed an replaces the application previously implemented as part of the Machine Learning Backend (uploader_app.py)