audioclass v0.1.1

This is the first release of audioclass, a Python library designed to simplify your audio classification projects. Here's what you can expect:

Key Features:

• Standardized Model Interface: Easily work with different audio classification models through a unified API. Currently supports:
  • BirdNET
  • Perch
• Flexible Data Handling:
  • Load audio recordings from files, directories, or pandas DataFrames.
  • Easily resample and preprocess audio data to match your model's requirements.
• Efficient Batch Processing:
  • Quickly process large audio datasets with SimpleIterator for sequential processing.
  • For larger datasets or parallel processing needs, explore TFDatasetIterator (requires TensorFlow).
• Versatile Output Formats:
  • Get predictions as easy-to-use ClipPrediction objects from the soundevent library.
  • Optionally obtain results as xarray Datasets for structured analysis.

Getting Started:

1. Install: pip install audioclass (Add [birdnet] or [perch] for the respective models)
2. Explore Examples: Check out our documentation for code examples and tutorials.

Looking Ahead:

This is just the beginning! We have exciting plans for future releases, including:

• More pre-trained models for various audio classification tasks
• Advanced preprocessing and postprocessing features
• Improved batch processing performance and flexibility
• Expanded documentation and examples

Feedback Welcome!

We'd love to hear your feedback on this initial release. Let us know what you think, what features you'd like to see, and any issues you encounter. Together, we can build a vibrant community around audioclass and make audio classification easier and more accessible for everyone.