ANIMAL-CLEAN – A Deep Denoising Toolkit for Animal-Independent Signal Enhancement

Barnhill A, Nöth E, Maier A, Bergler C (2024)

Publication Language: English

Publication Type: Conference contribution, Original article

Publication year: 2024

Conference Proceedings Title: Proceedings of the Annual Conference of the International Speech Communication Association

Event location: Kos, Greece GR

Abstract

Signal enhancement in bioacoustics can be of vital importance due to the fact that recordings are largely done in noise-heavy environments, in which anthrophonic, geophonic, or biophonic disturbances are myriad and can impede downstream analysis. Existing audio denoising techniques largely focus on human speech, whereas non-human animal vocalizations may not be compatible with frequency or temporal assumptions present in those methods. This work introduces ANIMAL--CLEAN, an animal-independent, Noise2Noise-based deep denoising toolkit utilizing a combination of signals from the target of interest as well as existing datasets used for human speech enhancement. This toolkit enables users to develop deep learning models capable of denoising bioacoustic signals with a wide range of frequencies and signal lengths, which, when used as a preprocessing step, improve results for downstream signal segmentation and classification, without the need for noise-free bioacoustic recordings.

Authors with CRIS profile

Alexander Barnhill Lehrstuhl für Informatik 5 (Mustererkennung) Elmar Nöth Professur für Informatik (Mustererkennung) Andreas Maier Lehrstuhl für Informatik 5 (Mustererkennung)

Involved external institutions

Ostbayerische Technische Hochschule (OTH) Amberg-Weiden DE Germany (DE)

How to cite

APA:

Barnhill, A., Nöth, E., Maier, A., & Bergler, C. (2024). ANIMAL-CLEAN – A Deep Denoising Toolkit for Animal-Independent Signal Enhancement. In Proceedings of the Annual Conference of the International Speech Communication Association. Kos, Greece, GR.

MLA:

Barnhill, Alexander, et al. "ANIMAL-CLEAN – A Deep Denoising Toolkit for Animal-Independent Signal Enhancement." Proceedings of the 25th Annual Conference of the International Speech Communication Association, INTERSPEECH 2024, Kos, Greece 2024.

BibTeX: Download