About

Tutorials written in Jupyter Notebooks to guide you through both accessing and using the data in the AWS pacific-sound registry.

Tutorials

Data Notebooks

• 2kHz Data ✨ Recommended first step to learn more about the low-frequency data
• 16kHz Data ✨ Recommended first step to learn more about the low-mid-frequency data
• 256kHz Data ✨ Recommended first step to learn more about the raw 10-minute data
• 256kHz to 2kHz Decimation ✨ Recommended to learn about how the raw 10-minute data is decimated from 256kHz to 2Khz

Research and Machine Learning Notebooks

• Blue Whales
  • Blue B call index 🐳 Study song occurrence using a signal processing method
  • Blue A call classification 🐳 Classify blue whale song A calls with a neural network model
• Fin Whales
  • Fin whale call index 🐳 Fin whale song occurrence using a signal processing method
• Humpback Whales
  • Humpback whale song detection 🐳 Detect humpback song with a neural network model
• Shipping Noise
  • Quantify shipping noise in the soundscape 🛳️ Apply international standards to measure shipping noise and its temporal variations.

Listen Up

• Full access to 16 kHz audio 🐬 Listen to example recordings of dolphins and whales, then pick any time to listen and explore.

Documentation

See pacific-sound for the official documentation.

Quick Start

Install in SageMaker

Install in Colab

Install in your computer

Python>=3.6.0 is required with the requirements.txt.

The recommended way to run the notebooks is using the Anaconda environment

git clone https://github.com/mbari-org/pacific-sound-notebooks
cd pacific-sound-notebooks

If using a Mac

brew install sox

If using Linux

apt-get install libsox-fmt-all libsox-dev

conda env create 
conda activate pacific-sound-notebooks
pip install ipykernel
python -m ipykernel install --user --name=pacific-sound-notebooks
jupyter notebook