The biodiversity crisis is maybe the most pressing issue of our time, which makes measuring the biodiversity an important precursor for improvement.
Biodiversity assessment is limited by data management and analysis tools. Prior work has approached this problem by exploring the space of tools that can aide workflows of biodiversity assessment. One of the available tools for bioacoustic data analysis is the Python library SciKit-MAAD, which provides high-level APIs for building data driven analysis workflows.
SciKit-MAAD includes functionality that aides a workflow of visualizing, filtering, binarizing and segmenting regions of interests in an audio file. However, the Python API requires its users to have programming expertise, which cannot be considered a given in taxonomy experts.
Figure 1. Source: scikit‐maad
In this work, we are providing an easy-to-use web-based implementation of the workflow proposed in the SciKit-MAAD publication. This solution should make this tool more approachable to taxonomists without programming knowledge.
Figure 2: Our application enables the user to record or upload an audio file of their own or select one from a list of predefined ones. After uploading, the audio file is automatically analysed with the functionality provided by SciKit-MAAD.
While the automatic data analysis functionality provided by SciKit-MAAD is replicated in our application, there is little interaction. The user can zoom and pan the plotted spectrogram in their web browser, but the automatically detected regions of interest cannot be edited. This can be the subject of future work and theses.
Links and Publications
ULLOA, Juan Sebastián, et al. scikit‐maad: An open‐source and modular toolbox for quantitative soundscape analysis in Python. Methods in Ecology and Evolution, 2021, 12. Jg., Nr. 12, S. 2334-2340.