Prediction of subjective assessments for a noise map using deep neural networks

Shota Kobayashi, Masanobu Abe, Sunao Hara

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper, we investigate a method of creating noise maps that take account of human senses. Physical measurements are not enough to design our living environment and we need to know subjective assessments. To predict subjective assessments from loudness values, we propose to use metadata related to where, who and what is recording. The proposed method is implemented using deep neural networks because these can naturally treat a variety of information types. First, we evaluated its performance in predicting five-point subjective loudness levels based on a combination of several features: location-specific, participant-specific, and sound-specific features. The proposed method achieved a 16.3 point increase compared with the baseline method. Next, we evaluated its performance based on noise map visualization results. The proposed noise maps were generated from the predicted subjective loudness level. Considering the differences between the two visualizations, the proposed method made fewer errors than the baseline method.

Original languageEnglish
Title of host publicationUbiComp/ISWC 2017 - Adjunct Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable Computers
PublisherAssociation for Computing Machinery, Inc
Pages113-116
Number of pages4
ISBN (Electronic)9781450351904
DOIs
Publication statusPublished - Sep 11 2017
Event2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and ACM International Symposium on Wearable Computers, UbiComp/ISWC 2017 - Maui, United States
Duration: Sep 11 2017Sep 15 2017

Other

Other2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and ACM International Symposium on Wearable Computers, UbiComp/ISWC 2017
CountryUnited States
CityMaui
Period9/11/179/15/17

Fingerprint

Visualization
Metadata
Acoustic waves
Deep neural networks

Keywords

  • Loudness
  • Map
  • Noise
  • Subjective
  • Visualization

ASJC Scopus subject areas

  • Software
  • Hardware and Architecture
  • Computer Networks and Communications

Cite this

Kobayashi, S., Abe, M., & Hara, S. (2017). Prediction of subjective assessments for a noise map using deep neural networks. In UbiComp/ISWC 2017 - Adjunct Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable Computers (pp. 113-116). Association for Computing Machinery, Inc. https://doi.org/10.1145/3123024.3123091

Prediction of subjective assessments for a noise map using deep neural networks. / Kobayashi, Shota; Abe, Masanobu; Hara, Sunao.

UbiComp/ISWC 2017 - Adjunct Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable Computers. Association for Computing Machinery, Inc, 2017. p. 113-116.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kobayashi, S, Abe, M & Hara, S 2017, Prediction of subjective assessments for a noise map using deep neural networks. in UbiComp/ISWC 2017 - Adjunct Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable Computers. Association for Computing Machinery, Inc, pp. 113-116, 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and ACM International Symposium on Wearable Computers, UbiComp/ISWC 2017, Maui, United States, 9/11/17. https://doi.org/10.1145/3123024.3123091
Kobayashi S, Abe M, Hara S. Prediction of subjective assessments for a noise map using deep neural networks. In UbiComp/ISWC 2017 - Adjunct Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable Computers. Association for Computing Machinery, Inc. 2017. p. 113-116 https://doi.org/10.1145/3123024.3123091
Kobayashi, Shota ; Abe, Masanobu ; Hara, Sunao. / Prediction of subjective assessments for a noise map using deep neural networks. UbiComp/ISWC 2017 - Adjunct Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable Computers. Association for Computing Machinery, Inc, 2017. pp. 113-116
@inproceedings{483c783906734f2286ff876625cf2a0f,
title = "Prediction of subjective assessments for a noise map using deep neural networks",
abstract = "In this paper, we investigate a method of creating noise maps that take account of human senses. Physical measurements are not enough to design our living environment and we need to know subjective assessments. To predict subjective assessments from loudness values, we propose to use metadata related to where, who and what is recording. The proposed method is implemented using deep neural networks because these can naturally treat a variety of information types. First, we evaluated its performance in predicting five-point subjective loudness levels based on a combination of several features: location-specific, participant-specific, and sound-specific features. The proposed method achieved a 16.3 point increase compared with the baseline method. Next, we evaluated its performance based on noise map visualization results. The proposed noise maps were generated from the predicted subjective loudness level. Considering the differences between the two visualizations, the proposed method made fewer errors than the baseline method.",
keywords = "Loudness, Map, Noise, Subjective, Visualization",
author = "Shota Kobayashi and Masanobu Abe and Sunao Hara",
year = "2017",
month = "9",
day = "11",
doi = "10.1145/3123024.3123091",
language = "English",
pages = "113--116",
booktitle = "UbiComp/ISWC 2017 - Adjunct Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable Computers",
publisher = "Association for Computing Machinery, Inc",

}

TY - GEN

T1 - Prediction of subjective assessments for a noise map using deep neural networks

AU - Kobayashi, Shota

AU - Abe, Masanobu

AU - Hara, Sunao

PY - 2017/9/11

Y1 - 2017/9/11

N2 - In this paper, we investigate a method of creating noise maps that take account of human senses. Physical measurements are not enough to design our living environment and we need to know subjective assessments. To predict subjective assessments from loudness values, we propose to use metadata related to where, who and what is recording. The proposed method is implemented using deep neural networks because these can naturally treat a variety of information types. First, we evaluated its performance in predicting five-point subjective loudness levels based on a combination of several features: location-specific, participant-specific, and sound-specific features. The proposed method achieved a 16.3 point increase compared with the baseline method. Next, we evaluated its performance based on noise map visualization results. The proposed noise maps were generated from the predicted subjective loudness level. Considering the differences between the two visualizations, the proposed method made fewer errors than the baseline method.

AB - In this paper, we investigate a method of creating noise maps that take account of human senses. Physical measurements are not enough to design our living environment and we need to know subjective assessments. To predict subjective assessments from loudness values, we propose to use metadata related to where, who and what is recording. The proposed method is implemented using deep neural networks because these can naturally treat a variety of information types. First, we evaluated its performance in predicting five-point subjective loudness levels based on a combination of several features: location-specific, participant-specific, and sound-specific features. The proposed method achieved a 16.3 point increase compared with the baseline method. Next, we evaluated its performance based on noise map visualization results. The proposed noise maps were generated from the predicted subjective loudness level. Considering the differences between the two visualizations, the proposed method made fewer errors than the baseline method.

KW - Loudness

KW - Map

KW - Noise

KW - Subjective

KW - Visualization

UR - http://www.scopus.com/inward/record.url?scp=85030863542&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85030863542&partnerID=8YFLogxK

U2 - 10.1145/3123024.3123091

DO - 10.1145/3123024.3123091

M3 - Conference contribution

SP - 113

EP - 116

BT - UbiComp/ISWC 2017 - Adjunct Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable Computers

PB - Association for Computing Machinery, Inc

ER -