Markov Chain GANS for bee bioacoustic signal synthesis
Loading...
Supplementary material
Other Title
Authors
Samarappuli, Kumudu
Ardekani, Iman
Ardekani, Iman
Author ORCID Profiles (clickable)
Degree
Grantor
Date
2025-07
Supervisors
Type
Conference Contribution - Paper in Published Proceedings
Ngā Upoko Tukutuku (Māori subject headings)
Keyword
honeybees (Apis mellifera)
audio data processing systems
beehive monitoring
generative adversarial networks (Computer networks)
AI in agriculture
audio data processing systems
beehive monitoring
generative adversarial networks (Computer networks)
AI in agriculture
ANZSRC Field of Research Code (2020)
Citation
Samarappuli, K. H., & Ardekani, I. (2025) Markov Chain GANS for bee bioacoustic signal synthesis. In S. Varastehpour & M. Shakiba (Eds.), Proceedings: AIOT Global Summit 2025: Economic Growth, 15–16 July (pp. 62 66). ePress, Unitec. https://doi.org/10.34074/proc.250112
Abstract
One of the key challenges in processing bioacoustic signals through machine learning is the lack of high-quality and sufficient data. Research suggests that using synthetically generated data is a promising solution to this challenge, because it offers a clean and balanced dataset. As there are existing approaches to generating synthetic data, this study analyses them specifically considering the bioacoustics domain and proposes using Markov chain generative adversarial networks to address that challenge. The study uses the generative power of WaveGAN and Markov chain process to improve the realism of generated data. This research aims to fill a critical gap in synthetic bioacoustic signal generation and also provide effective support for bee-related activities and future research in bioacoustics
Publisher
Unitec ePress
Permanent link
Link to ePress publication
DOI
https://doi.org/10.34074/proc.250112
Copyright holder
Authors
Copyright notice
CC BY-NC Attribution-NonCommercial 4.0 International