OSEkit is an open source python package dedicated to the management and analysis of data in underwater passive acoustics.
OSEkit is a Python open source package designed for effortless manipulation of audio data.
It is primarily designed for working on underwater passive acoustics data, but can be used on any audio data.
OSEkit provides:
- Seamless, timestamp-based access across multiple files (e.g., pick a 3-minutes long segment spannign over two 1.5-minutes long files without manual file picking)
- Preprocessing utilities: resampling, normalization...
- Spectral analysis routines: compute power spectra, spectrograms, LTAS...
- A high-level API designed to work with APLOSE, our web-base spectrogram annotation tool
OSEkit treats your audio dataset as a continuous timeline, whatever the raw files configuration. You request audio by time intervals and the package fetches from the raw files behind the scenes: no manual I/O juggling required.
All details to start using our toolkit and make the most out of it are given in our documentation.
This software is distributed under the MIT license.
You are free to use, modify, and redistribute it, provided that you retain the copyright notices and license text, in accordance with the terms of this license.
See the LICENSE file for the full text.
Some third-party libraries are used in this project. Details of their licenses are available in the repository dependency graph.
This software must not be sold.
If you use this software in scientific work, please cite it in accordance with the CITATION file.
This software is provided for research and teaching purposes. It is provided “as is,” without warranty. The authors and the institution cannot be held responsible for any damage resulting from its use.
© OSmOSE team, 2023-present