Honey Bee Queen Presence Detection from Audio Field Recordings using Summarized Spectrogram and Convolutional Neural Networks - Archive ouverte HAL Access content directly
Conference Papers Year : 2021

Honey Bee Queen Presence Detection from Audio Field Recordings using Summarized Spectrogram and Convolutional Neural Networks

(1) , (1) , (2) , (2)
1
2

Abstract

The present work proposes a simple supervised method based on a downsampled time-frequency representation of the input audio signal for detecting the presence of the queen in a beehive from noisy field recordings. Our proposed technique computes a "summarized-spectrogram" of the signal that is used as the input of a deep convolutional neural network. This approach has the advantage of reducing the dimension of the input layer and the computational cost while obtaining better classification results with the same deep neural architecture. Our comparative evaluation based on a cross-validation beehive-independent methodology shows a maximal accuracy of 96% using the proposed approach applied on the evaluation dataset. This corresponds to a significant improvement of the prediction accuracy in comparison to several state-of-the-art approaches reported by the literature. Baseline methods such as MFCC, constant-Q transform and classical STFT combined with a CNN fail to generalize the prediction of the queen presence in an unknown beehive and obtain a maximal accuracy of 55% in our experiments.
Fichier principal
Vignette du fichier
article.pdf (4.15 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03439646 , version 1 (22-11-2021)

Identifiers

Cite

Agnieszka Orlowska, Dominique Fourer, Jean-Paul Gavini, Dominique Cassou-Ribehart. Honey Bee Queen Presence Detection from Audio Field Recordings using Summarized Spectrogram and Convolutional Neural Networks. 21st International Conference on Intelligent Systems Design and Applications (ISDA 2021), Dec 2021, Seattle, WA, (World Wide Web), United States. pp.83--92, ⟨10.1007/978-3-030-96308-8_8⟩. ⟨hal-03439646⟩
175 View
37 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More