CASE (Cluster & Analyse Sound Events)
A tool to evaluate bioacoustic studies easily and automatically
現在この提出コンテンツをフォロー中です。
- フォローしているコンテンツ フィードに更新が表示されます。
- コミュニケーション基本設定に応じて電子メールを受け取ることができます
By means of this tool it is possible to extract up to 44 features of an animal sound in different ways. In addition, the tool includes various clustering algorithms (community detection; affinity propagation; HDBSCAN and fuzzy clustering) as well as algorithms to detect similarities between signals (k-nearest-neighbor; jaccard; dynamic-time-warping) in order to effectively classify animal sounds. The tool uses a graphical user interface to allow the user to work with the software as easily and intuitively as possible.
The following algorithms were adopted:
Affinity Propagation:
Kaijun Wang (2021). Adaptive Affinity Propagation clustering (https://www.mathworks.com/matlabcentral/fileexchange/18244-adaptive-affinity-propagation-clustering), MATLAB Central File Exchange. Retrieved March 1, 2021.
Community Detection:
Athanasios Kehagias (2021). Community Detection Toolbox (https://www.mathworks.com/matlabcentral/fileexchange/45867-community-detection-toolbox), MATLAB Central File Exchange. Retrieved March 1, 2021.
HDBSCAN:
Jordan Sorokin (2021). Jorsorokin/HDBSCAN (https://github.com/Jorsorokin/HDBSCAN), GitHub. Retrieved March 1, 2021.
NMI:
Mo Chen (2021). Normalized Mutual Information (https://www.mathworks.com/matlabcentral/fileexchange/29047-normalized-mutual-information), MATLAB Central File Exchange. Retrieved March 1, 2021.
t-SNE:
Laurens van der Maaten & Geoffrey Hinton
https://lvdmaaten.github.io/tsne/
引用
Schneider, Sebastian, et al. “Introducing the Software CASE (Cluster and Analyze Sound Events) by Comparing Different Clustering Methods and Audio Transformation Techniques Using Animal Vocalizations.” Animals, vol. 12, no. 16, MDPI AG, Aug. 2022, p. 2020, doi:10.3390/ani12162020.
