HRTF personalization based on artificial neural network in individual virtual auditory space

Hu, Hongmei; Lin, Zhenzhi; Ma, Hao; Wu, Zhenyang

doi:10.1016/j.apacoust.2007.05.007

Cited by 93 publications

(68 citation statements)

References 10 publications

(15 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A three-layer BPNN for nonlinear regression was compared with the proposed approach. According to [6], 8 anthropometric parameters including x 1 , x 3 , x 12 , d 1 , d 3 − d 6 were used as the inputs. The outputs were 12 dimensional weight vectors obtained with PCA.…”

Section: Resultsmentioning

confidence: 99%

“…However, because of the presence of the complex scattering of the incident sound by the head, shoulder, torso and pinna, it is necessary to adopt nonlinear regression methods to customize individual HRIR. Subsequently they presented a nonlinear approach based on a three-layer back-propagation neural network (BPNN) [6] . The inputs of the neural network are some anthropometric parameters selected by correlation analysis and the outputs are the principal components together with the interaural time difference (ITD) of HRTF.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Modeling personalized head-related impulse response using support vector regression

Huang

Fang

2009

J. Shanghai Univ.(Engl. Ed.)

View full text Add to dashboard Cite

A new customization approach based on support vector regression (SVR) is proposed to obtain individual headrelated impulse response (HRIR) without complex measurement and special equipment. Principal component analysis (PCA) is first applied to obtain a few principal components and corresponding weight vectors correlated with individual anthropometric parameters. Then the weight vectors act as output of the nonlinear regression model. Some measured anthropometric parameters are selected as input of the model according to the correlation coefficients between the parameters and the weight vectors. After the regression model is learned from the training data, the individual HRIR can be predicted based on the measured anthropometric parameters. Compared with a back-propagation neural network (BPNN) for nonlinear regression, better generalization and prediction performance for small training samples can be obtained using the proposed PCA-SVR algorithm.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Modeling personalized head-related impulse response using support vector regression

Huang

Fang

2009

J. Shanghai Univ.(Engl. Ed.)

View full text Add to dashboard Cite

show abstract

“…Due to the fact that HRTF individualization is strongly related to the anthropometry of a person, methods have been proposed for HRTF personalization by choosing a small set of anthropometry features with a pre-trained model [49][50][51][52][53][54]. The training was established based on a direct linear or nonlinear relationship between the anthropometric data and the HRTFs, where the first step is to reduce the HRTF data dimensionality.…”

Section: Individualized Hrtfmentioning

confidence: 99%

Surround by Sound: A Review of Spatial Audio Recording and Reproduction

et al. 2017

View full text Add to dashboard Cite

Abstract:In this article, a systematic overview of various recording and reproduction techniques for spatial audio is presented. While binaural recording and rendering is designed to resemble the human two-ear auditory system and reproduce sounds specifically for a listener's two ears, soundfield recording and reproduction using a large number of microphones and loudspeakers replicate an acoustic scene within a region. These two fundamentally different types of techniques are discussed in the paper. A recent popular area, multi-zone reproduction, is also briefly reviewed in the paper. The paper is concluded with a discussion of the current state of the field and open problems.

show abstract

“…Low-dimensional PCA representations of HRTFs are often used as targets for regression/interpolation and personalization from predictors such as anthropometry [3,4]. While PCA captures maximal variance along linear bases, non-linear relationships that are visible in HRTFs such as shifted spectral cues (notches/peaks) and smoothness assumptions along frequency are not represented in the versions synthesized using the linear principal components.…”

Section: Introductionmentioning

confidence: 99%

“…While this may preserve some of the common spectral features in the low frequency ranges, many subject-specific features due to anthropometric variations are lost, leading to localization errors [5]. Many works have sought individualized HRTFs by learning their relationships to subject's anthropometry [6,3,4]. A recent idea achieves some progress by granting listeners full-access to nonindividualized HRTF PCA weights for interactive tuning [7]; while no anthropometric measurements are needed, the user must learn to tune PCA weights w.r.t.…”

Section: Introductionmentioning

confidence: 99%

Virtual autoencoder based recommendation system for individualizing head-related transfer functions

Luo

Zotkin

Duraiswami

2013

2013 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics

View full text Add to dashboard Cite

We propose a virtual autoencoder based recommendation system for learning a user's Head-related Transfer Functions (HRTFs) without subjecting a listener to impulse response or anthropometric measurements. Autoencoder neural-networks generalize principal component analysis (PCA) and learn non-linear feature spaces that supports both out-of-sample embedding and reconstruction; this may be applied to developing a more expressive low-dimensional HRTF representation. One application is to individualize HRTFs by tuning along the autoencoder feature spaces. We demonstrate this new approach by developing a virtual (black-box) user that can localize sound from query HRTFs reconstructed from those spaces. Standard optimization methods tune the autoencoder features based on the virtual user's feedback. Experiments with CIPIC HRTFs show that the virtual user can localize along out-of-sample directions and that optimization in the autoencoder feature space improves upon initial non-individualized HRTFs. Other applications of the representation are also discussed.

show abstract

HRTF personalization based on artificial neural network in individual virtual auditory space

Cited by 93 publications

References 10 publications

Modeling personalized head-related impulse response using support vector regression

Modeling personalized head-related impulse response using support vector regression

Surround by Sound: A Review of Spatial Audio Recording and Reproduction

Virtual autoencoder based recommendation system for individualizing head-related transfer functions

Contact Info

Product

Resources

About