View in EDS

The Affective Audio Dataset (AAD) for Non-Musical, Non-Vocalized, Audio Emotion Research

Saved in:
Bibliographic Details
Title: The Affective Audio Dataset (AAD) for Non-Musical, Non-Vocalized, Audio Emotion Research
Authors: Harrison Ridley, Stuart Cunningham, John Darby, John Henry, Richard Stocker
Contributors: University of Chester, Manchester Metropolitan University
Source: IEEE Transactions on Affective Computing. 16:394-404
Publisher Information: Institute of Electrical and Electronics Engineers (IEEE), 2025.
Publication Year: 2025
Subject Terms: 4603 Computer vision and multimedia computation, Numerical models, 4602 Artificial intelligence, 0806 Information Systems, 1702 Cognitive Sciences, Data collection, 0801 Artificial Intelligence and Image Processing, Affective computing, Emotion recognition, 4608 Human-centred computing, Music
Description: © 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. ; The Affective Audio Dataset (AAD) is a new and novel dataset of non-musical, non-anthropomorphic sounds intended for use in affective research. Sounds are annotated for their affective qualities by sets of human participants. The dataset was created in response to a lack of suitable datasets within the domain of audio emotion recognition. A total of 780 sounds are selected from the BBC Sounds Library. Participants are recruited online and asked to rate a subset of sounds based on how they make them feel. Each sound is rated for arousal and valence. It was found that while evenly distributed, there was bias towards the low-valence, high-arousal quadrant, and displayed a greater range of ratings in comparison to others. The AAD is compared with existing datasets to check its consistency and validity, with differences in data collection methods and intended use-cases highlighted. Using a subset of the data, the online ratings were validated against an in-person data collection experiment with findings strongly correlating. The AAD is used to train a basic affect-prediction model and results are discussed. Uses of this dataset include, human-emotion research, cultural studies, other affect-based research, and industry use such as audio post-production, gaming, and user-interface design. ; Unfunded ; AAM added 13/08/2024.
Document Type: Article
ISSN: 2371-9850
DOI: 10.1109/taffc.2024.3437153
Access URL: https://e-space.mmu.ac.uk/635285/
Rights: IEEE Copyright
CC BY NC ND
Accession Number: edsair.doi.dedup.....8e071915db9f9c290c4d74c424da037f
Database: OpenAIRE
Description
Abstract:© 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. ; The Affective Audio Dataset (AAD) is a new and novel dataset of non-musical, non-anthropomorphic sounds intended for use in affective research. Sounds are annotated for their affective qualities by sets of human participants. The dataset was created in response to a lack of suitable datasets within the domain of audio emotion recognition. A total of 780 sounds are selected from the BBC Sounds Library. Participants are recruited online and asked to rate a subset of sounds based on how they make them feel. Each sound is rated for arousal and valence. It was found that while evenly distributed, there was bias towards the low-valence, high-arousal quadrant, and displayed a greater range of ratings in comparison to others. The AAD is compared with existing datasets to check its consistency and validity, with differences in data collection methods and intended use-cases highlighted. Using a subset of the data, the online ratings were validated against an in-person data collection experiment with findings strongly correlating. The AAD is used to train a basic affect-prediction model and results are discussed. Uses of this dataset include, human-emotion research, cultural studies, other affect-based research, and industry use such as audio post-production, gaming, and user-interface design. ; Unfunded ; AAM added 13/08/2024.
ISSN:23719850
DOI:10.1109/taffc.2024.3437153