Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/164983
Title: A benchmark comparison of perceptual models for soundscapes on a large-scale augmented soundscape dataset
Authors: Ooi, Kenneth
Watcharasupat, Karn N.
Lam, Bhan
Ong, Zhen-Ting
Gan, Woon-Seng
Keywords: Engineering::Electrical and electronic engineering
Issue Date: 2022
Source: Ooi, K., Watcharasupat, K. N., Lam, B., Ong, Z. & Gan, W. (2022). A benchmark comparison of perceptual models for soundscapes on a large-scale augmented soundscape dataset. 24th International Congress on Acoustics (ICA 2022).
Project: COT-V4-2020-1 
Conference: 24th International Congress on Acoustics (ICA 2022)
Abstract: Psychoacoustic indicators and spectrogram-based features are standard inputs to perceptual models for soundscape analysis. However, existing models in the soundscape literature are trained on different collections of input parameters and frequently mutually exclusive datasets, which complicates fair comparisons of model performance and generalizability to unseen data and soundscapes. Hence, we use the ARAUS dataset, a large-scale, publicly-available dataset of 25,440 responses to unique augmented soundscapes, as a common benchmark for comparison of the relative performance of a selection of input and model types used in previous soundscape studies, as well as deep neural network architectures commonly used for other acoustic tasks. The different model types were used in a regression task to predict “Eventfulness” ratings (as defined in ISO/TS 12913-3) given by participants as responses to the augmented soundscapes. Subsequently, we compared their performance as classification models for the classification of soundscapes into the quadrants generated by the Pleasantness-Eventfulness axes of the ISO/TS 12913-3 circumplex model. The five-fold cross-validation set of 25,200 responses and an independent test set of 240 responses making up the ARAUS dataset was used to facilitate unbiased comparisons.
URI: https://hdl.handle.net/10356/164983
URL: https://www.ica2022korea.org/data/Proceedings_A14.pdf
https://ica2022korea.org/
DOI (Related Dataset): 10.21979/N9/9OTEVX
Schools: School of Electrical and Electronic Engineering 
Research Centres: Digital Signal Processing Laboratory 
Rights: © 2022 The Author(s). Published by Acoustical Society of Kore. All rights reserved. This paper was published in Proceedings of the 24th International Congress on Acoustics (ICA 2022) and is made available with permission of The Author(s).
Fulltext Permission: open
Fulltext Availability: With Fulltext
Appears in Collections:EEE Conference Papers

Files in This Item:
File Description SizeFormat 
ICA_2022_template_final_ABS-0603v2.pdf487.64 kBAdobe PDFThumbnail
View/Open

Page view(s)

173
Updated on Jun 18, 2024

Download(s) 50

67
Updated on Jun 18, 2024

Google ScholarTM

Check

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.