Using the BirdNet Pipeline on AML

Scheduled Pipeline Execution

The primary usage of the BirdNet pipeline is through a scheduled endpoint, which triggers automatically on set days. This scheduled deployment ensures that the pipeline processes new audio data consistently without requiring manual intervention.

Manual Pipeline Execution

Although the pipeline is designed to run automatically, you can trigger it manually if needed:

1. Trigger the Pipeline: Navigate to the Pipelines section in AML Studio.
2. Run the Pipeline: Select the BirdNet pipeline endpoint and click Submit.

Pipeline Phases

• Phase 1: Runs audio analysis using the analyze.py script to detect bird species.
• Phase 2: Processes and organizes predictions using the segments.py script.
• Phase 3: Archives the original audio files for storage.

Accessing Results

• After execution, processed results are stored in the bronze_audio_data datastore.
• You can access detailed logs and output files directly in AML Studio.