Training data set and python code for a classifier to detect lion roaring in ACC data - Supplementary to the paper: Did U hear that? Working with mixed behaviours when classifying animal behaviour from acceleration data using a U-Net

This data set contains acceleration data from five lions in Etosha National Park from 2022 and 2023. These data were used to train a U-Net that recognises roaring behhaviour in the ACC data. The ACC data was recorded continously at 32Hz with -4g to 4g senitivity. Part of this data set are also the behaviour labels with timestamps for the acceleration data as well as the python scripts to process the data as well as train and evaluate the model. The trained model is also included in a file.