1 code implementation • 29 Nov 2021 • Yuren Sun, Tatiana Midori Maeda, Claudia Solis-Lemus, Daniel Pimentel-Alarcon, Zuzana Burivalova
Using soundscapes from a tropical forest in Borneo and a Convolutional Neural Network model (CNN) created with transfer learning, we investigate i) the minimum viable training data set size for accurate prediction of call types ('sonotypes'), and ii) the extent to which data augmentation can overcome the issue of small training data sets.