[en] Woodpeckers calls are readily recognizable on spectrograms and this opens the door for their identification from images through a convolutional Deep Neural Network (DNN). We built a dataset of 12154 images, half woodpecker calls (9 classes) and half noise, from Xeno-Canto and private recordings. We experimented with two approaches: a) training a small net (2 convolutional layers, 2 dense layers) from scratch using the theano framework and b) re-training legacy image nets (over 150 layers) using Pytorch. The larger nets successfully differentiate the calls from noise and identify them with an accuracy greater than 90%.
Disciplines :
Computer science Zoology Physics
Author, co-author :
Florentin, Juliette ; Université de Mons > Faculté Polytechnique > Mécanique rationnelle, Dynamique et Vibrations
Verlinden, Olivier ; Université de Mons > Faculté Polytechnique > Service de Mécanique rationnelle, Dynamique et Vibrations
Dutoit, Thierry ; Université de Mons > Faculté Polytechnique > Service Information, Signal et Intelligence artificielle
Laraba, Sohaib ; Université de Mons > Faculté Polytechnique > Service Information, Signal et Intelligence artificielle
Language :
English
Title :
Recognition of woodpecker calls using a convolutional deep neural network
Publication date :
05 March 2019
Number of pages :
1
Event name :
Mardi des Chercheurs 2019 (MdC2019)
Event place :
Mons, Belgium
Event date :
2019
Research unit :
F703 - Mécanique rationnelle, Dynamique et Vibrations