Chimpanzee Drumming Detector

This is a python program for detecting chimpanzee drumming in long-term rainforest recordings.

The associated publication can be read at:

• https://arxiv.org/abs/2105.12502
• https://doi.org/10.1016/j.ecoinf.2021.101423

This program only contains the test pipeline, not the training pipeline.

Installation

Either install the latest stable release from PyPI

pip install ChimpDrummingDetector

Alternatively, you can install this package through to code on this repository. Simply download this repository as a zip, and then do:

cd path/to/downloaded/zip
unzip ChimpDrummingDetector.zip
cd ChimpDrummingDetector
pip install ChimpDrummingDetector

Usage

In your python code, do

import chimp_detector

path_to_wave_to_analyze = path/to/wave.wav 
detection_result = chimp_detector.detect_drumming(path_to_wave_to_analyze)

• Input: A wave file. Might be mono or stereo
• Output: A pandas dataframe with following columns
  • timepoints_in_seconds [float]: The analyzed timepoints
  • drumming_probability [float]: Probability of chimpanzee drumming in range [0,1] for the respective timepoint
  • drumming_binarized [boolean]: Binarization of probability with threshold 0.5, where TRUE means "there is drumming in this frame".

Dependencies

This Packages requires:

• Python version 3.8
• librosa==0.8.0
• tensorflow==2.3.1
• pandas==1.1.2
• numpy==1.18.5

Citation

Please cite this paper in your publication if you use this program for your research.

@article{anders2021compensating,
  title={Compensating class imbalance for acoustic chimpanzee detection with convolutional recurrent neural networks},
  author={Anders, Franz and Kalan, Ammie K and K{\"u}hl, Hjalmar S and Fuchs, Mirco},
  journal={Ecological Informatics},
  volume={65},
  pages={101423},
  year={2021},
  publisher={Elsevier}
  }