Individual HRTF Prediction Based on Anthropometric Data and Multi-Stage Model

Yinliang Qiu, Zhiyu Li, Jing Wang*

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

Getting individual head related transfer function (HRTF) is an important step in rendering binaural immersive audio. Individual HRTF can provide a more realistic experience than general HRTF. For more accurate prediction results, we propose a multi-stage model perform individual HRTF prediction based on anthropometric data. This model can combine global and local features through different stages. In the first stage, light gradient boosting machine(LightGBM) is chosen as decision tress model to predict HRTF according to anthropometric data and different angels. In the second stage, Transformer encoder is chosen to learn the global information between different frequency points. According to the experimental results, the effect of using a multi-stage model is better than that of a single model. The spectral distortion of the results predicted by our model is smaller, which can illustrate the effectiveness of our model.

Original languageEnglish
Title of host publicationProceedings - 2023 IEEE International Conference on Multimedia and Expo Workshops, ICMEW 2023
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages314-319
Number of pages6
ISBN (Electronic)9798350313154
DOIs
Publication statusPublished - 2023
Event2023 IEEE International Conference on Multimedia and Expo Workshops, ICMEW 2023 - Brisbane, Australia
Duration: 10 Jul 202314 Jul 2023

Publication series

NameProceedings - 2023 IEEE International Conference on Multimedia and Expo Workshops, ICMEW 2023

Conference

Conference2023 IEEE International Conference on Multimedia and Expo Workshops, ICMEW 2023
Country/TerritoryAustralia
CityBrisbane
Period10/07/2314/07/23

Keywords

  • Individual HRTF
  • LightGBM
  • Transformer encoder
  • multi-stage model

Fingerprint

Dive into the research topics of 'Individual HRTF Prediction Based on Anthropometric Data and Multi-Stage Model'. Together they form a unique fingerprint.

Cite this