Binaural Sound Source Localization based on Neural Networks in Mismatched HRTF Condition

Kai Qian; Jing Wang; Wenjing Yang; Miao Liu

doi:10.1145/3532213.3532223

Binaural Sound Source Localization based on Neural Networks in Mismatched HRTF Condition

Kai Qian, Jing Wang^*, Wenjing Yang, Miao Liu

^*Corresponding author for this work

School of Information and Electronics

Beijing Institute of Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Citation (Scopus)

Abstract

Binaural sound source localization is a field with wide applications, such as virtual sound localization in VR or speech enhancement, which has drawn many researchers' attention. However, the mismatched HRTF condition is a severe problem, which has been ignored in most of the previous researches. In this paper, an experiment is firstly conducted to prove the negative influence of HRTF individualization on binaural localization performance. In face with this problem, an improved localization method is proposed in this paper. Both DNN and CNN are used in this method and the performance is compared. Then, another experiment is also operated to prove the efficiency of this method, providing ideas for subsequent iterations.

Original language	English
Title of host publication	ICCAI 2022 - Proceedings of 2022 8th International Conference on Computing and Artificial Intelligence
Publisher	Association for Computing Machinery
Pages	62-67
Number of pages	6
ISBN (Electronic)	9781450396110
DOIs	https://doi.org/10.1145/3532213.3532223
Publication status	Published - 18 Mar 2022
Event	8th International Conference on Computing and Artificial Intelligence, ICCAI 2022 - Virtual, Online, China Duration: 18 Mar 2022 → 21 Mar 2022

Publication series

Name	ACM International Conference Proceeding Series

Conference

Conference	8th International Conference on Computing and Artificial Intelligence, ICCAI 2022
Country/Territory	China
City	Virtual, Online
Period	18/03/22 → 21/03/22

Keywords

Binaural localization
HRTF
Neural network

Access to Document

10.1145/3532213.3532223

Cite this

Qian, K., Wang, J., Yang, W., & Liu, M. (2022). Binaural Sound Source Localization based on Neural Networks in Mismatched HRTF Condition. In ICCAI 2022 - Proceedings of 2022 8th International Conference on Computing and Artificial Intelligence (pp. 62-67). (ACM International Conference Proceeding Series). Association for Computing Machinery. https://doi.org/10.1145/3532213.3532223

@inproceedings{f47d04bcd3c34b03b9975fd29cbdc1d1,

title = "Binaural Sound Source Localization based on Neural Networks in Mismatched HRTF Condition",

abstract = "Binaural sound source localization is a field with wide applications, such as virtual sound localization in VR or speech enhancement, which has drawn many researchers' attention. However, the mismatched HRTF condition is a severe problem, which has been ignored in most of the previous researches. In this paper, an experiment is firstly conducted to prove the negative influence of HRTF individualization on binaural localization performance. In face with this problem, an improved localization method is proposed in this paper. Both DNN and CNN are used in this method and the performance is compared. Then, another experiment is also operated to prove the efficiency of this method, providing ideas for subsequent iterations.",

keywords = "Binaural localization, HRTF, Neural network",

author = "Kai Qian and Jing Wang and Wenjing Yang and Miao Liu",

note = "Publisher Copyright: {\textcopyright} 2022 ACM.; 8th International Conference on Computing and Artificial Intelligence, ICCAI 2022 ; Conference date: 18-03-2022 Through 21-03-2022",

year = "2022",

month = mar,

day = "18",

doi = "10.1145/3532213.3532223",

language = "English",

series = "ACM International Conference Proceeding Series",

publisher = "Association for Computing Machinery",

pages = "62--67",

booktitle = "ICCAI 2022 - Proceedings of 2022 8th International Conference on Computing and Artificial Intelligence",

}

Qian, K, Wang, J, Yang, W & Liu, M 2022, Binaural Sound Source Localization based on Neural Networks in Mismatched HRTF Condition. in ICCAI 2022 - Proceedings of 2022 8th International Conference on Computing and Artificial Intelligence. ACM International Conference Proceeding Series, Association for Computing Machinery, pp. 62-67, 8th International Conference on Computing and Artificial Intelligence, ICCAI 2022, Virtual, Online, China, 18/03/22. https://doi.org/10.1145/3532213.3532223

Binaural Sound Source Localization based on Neural Networks in Mismatched HRTF Condition. / Qian, Kai; Wang, Jing; Yang, Wenjing et al.
ICCAI 2022 - Proceedings of 2022 8th International Conference on Computing and Artificial Intelligence. Association for Computing Machinery, 2022. p. 62-67 (ACM International Conference Proceeding Series).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Binaural Sound Source Localization based on Neural Networks in Mismatched HRTF Condition

AU - Qian, Kai

AU - Wang, Jing

AU - Yang, Wenjing

AU - Liu, Miao

PY - 2022/3/18

Y1 - 2022/3/18

N2 - Binaural sound source localization is a field with wide applications, such as virtual sound localization in VR or speech enhancement, which has drawn many researchers' attention. However, the mismatched HRTF condition is a severe problem, which has been ignored in most of the previous researches. In this paper, an experiment is firstly conducted to prove the negative influence of HRTF individualization on binaural localization performance. In face with this problem, an improved localization method is proposed in this paper. Both DNN and CNN are used in this method and the performance is compared. Then, another experiment is also operated to prove the efficiency of this method, providing ideas for subsequent iterations.

AB - Binaural sound source localization is a field with wide applications, such as virtual sound localization in VR or speech enhancement, which has drawn many researchers' attention. However, the mismatched HRTF condition is a severe problem, which has been ignored in most of the previous researches. In this paper, an experiment is firstly conducted to prove the negative influence of HRTF individualization on binaural localization performance. In face with this problem, an improved localization method is proposed in this paper. Both DNN and CNN are used in this method and the performance is compared. Then, another experiment is also operated to prove the efficiency of this method, providing ideas for subsequent iterations.

KW - Binaural localization

KW - HRTF

KW - Neural network

UR - http://www.scopus.com/inward/record.url?scp=85134383173&partnerID=8YFLogxK

U2 - 10.1145/3532213.3532223

DO - 10.1145/3532213.3532223

M3 - Conference contribution

AN - SCOPUS:85134383173

T3 - ACM International Conference Proceeding Series

SP - 62

EP - 67

BT - ICCAI 2022 - Proceedings of 2022 8th International Conference on Computing and Artificial Intelligence

PB - Association for Computing Machinery

T2 - 8th International Conference on Computing and Artificial Intelligence, ICCAI 2022

Y2 - 18 March 2022 through 21 March 2022

ER -

Binaural Sound Source Localization based on Neural Networks in Mismatched HRTF Condition

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this